TROPHIC INTERRELATIONSHIPS} LIFE-HISTORIES AND TAXONOMY OF SOME INVERTEBRATES ASSOCIATED WITH AQUATIC MACROPHYTES IN LAKE GRASMERE

A thesis

presented for the degree of

Doctor of Philosophy in Zoology

in the

University of Canterbury,

Christchurch, New Zealand

by

JOHN DOUGLAS STARK

1981 i CONTENTS

Page

List of Tables v List of Plates vii List of Figures vii Abstract ix

CHAPTER GENERAL INTRODUCTION 1

I I STUDY AREA 3 2.1 INTRODUCTION 3 2.1.1 Location, Formation and Catchment of Lake Grasmere 3 2.2 CLIMATE 4 2 .3 LAKE GRASMERE 5 2.3.1 Physical Features 5 2.3.2 Chemical and Biological Features 6

III QUANTITATIVE SAMPLING PROGRAM 11 3.1 AIMS OF THE QUANTITATIVE SAMPLING PROGRAM 11 3.2 METHODS 12 3.2.1 Sampling Methods 12 3.2.2 pilot Survey and Analysis of Sample Variabili ty 18 3.2.3 The Main Quantitative Sampling Program 22 (a) Field procedure 22 (b) Laboratory procedure 24 3.3 RESULTS AND DISCUSSION 24 3.3.1 Sampling Variability 24 3.3.2 Community Species Composition 28 3.3.3 Known Habitat Requirements of Some Freshwater Invertebrates 34 3.3.4 Invertebrate Community Relationships 39 (1) Invertebrate communities on different macrophytes 39 (2) Invertebrate communities at different depths 43 (3) site groups and species groups 45 (a) Introduction 45 (b) Data processing 46 (c) si te groups 48 (d) Species groups 52 3.3,5 Seasonal Changes in Invertebrate Communities at the Site Groups 63 (1) Introduction 63 (2) Procedure 63 (a) Species diversity 63 (b) community similarity indices 64 (3) Results 65 (a) Species diversity 65 (b) Community similarity 67 ii

Chapter Page

3.3.6 Seasonal Changes in the Abundance of 70 Major Taxa at the site Groups (1) site group A 71 (2) Site group B 72 (3) Site group C 74 (4) Site group D 76 3.4 CONCLUSIONS 77 3.4.1 The Composition of Invertebrate Communities at the Site Groups in Terms of Species Groups and Major Taxa 77 3.4.2 Comparison of Species Diversity and Invertebrate Density at Different site Groups 83 3.4.3 Community Similarity 84

IV TROPHIC INTERRELATIONSHIPS OF SOME MACROPHYTE-ASSOCIATED INVERTEBRATES 87

4.1 INTRODUCTION 87 4.2 METHODS 89 4.2.1 Faecal Analysis 89 (1) Field procedure 89 (2) Laboratory procedure 91 4.2.2 Periphyton Analysis 95 4.3 RESULTS AND DISCUSSION 96 4.3.1 Introduction 96 4.3.2 Periphyton Analyses 97 4.3.3 Invertebrate Faecal Analyses 98 (1) Paroxyethira hendersoni and 98 P. tillyardi (2) Hudsonema amabilis 100 (3) Triplectides cephalotes 104 (4) Nymphula nitens 108 (5) Xanthocnemis zealandica 111 (6) Potamopyrgus antipodarum 115 (7) Observations on the gut contents of Chironomidae 118

V LIFE-HISTORY INFORMATION ON SELECTED 121 INSECTS

5.1 INTRODUCTION 121 5.2 METHODS 122 5.3 RESULTS AND DISCUSSION 124 5.3.1 Hudsonema amabilis 124 5.3.2 Triplectides cephalotes 126 5.3.3 pycnocentrodes aureola 127 5.3.4 Oecetis unicolor 129 5.3.5 Oxyethira albiceps, paroxyethira 130 hendersoni and P. tillyardi 5.3.6 Nymphula nitens 131 5.3.7 Xanthocnemis zealandica 133 iii

Chapter Page

VI TAXONOMY OF NEW ZEALAND HYDROPTILIDAE 137 (TRICHOPTERA) AND CHIRONOMIDAE (DI RA)

6.1 INTRODUCTION 137 6.2 TAXONOMY OF THE LARVAE OF NEW ZEALAND HYDROPTILIDAE 137 6.2.1 Key to the Larval Hydroptilidae of New Zealand 139 6.2.2 Distribution of the New Zealand species of Hydroptilidae 140 6.2.3 Description of the larva of paroxyethira tillyardi 144 6.3 TAXONOMY OF NEW ZEALAND CHIRONOMIDAE 147 6.3.1 Key to Larval Chironomidae of New Zealand 149 6.3.2 Key to Adult Male Chironomidae of New Zealand 160 6.3.3 Description of the Adult Male of Eukiefferiella sp. 171 6.3.4 Chironomid Larvae and from Lake Grasmere 173 (1) Tanypodinae 175 (2) Podonominae 177 (3) Orthocladiinae 177 (4) Chironominae 180

VII GENERAL DISCUSSION 181 ACKNOWLEDGMENTS 193 REF ENCES 195 APPENDICES

1 Full collection records of macrophyte-associated invertebrates collected during the pilot survey, 219 Lake Grasmere (14 April 1976) 2 Full collection records of macrophyte-associated invertebrates collected during the main 220-232 quantitative sampling program (September 1976 - October 1977) (Appendices 2.1- 2.13) 3 Chironomid larvae and pupae collected during the quantitative sampling program (September 1976 - 233-236 October 1977). 4 Indices of precision (D) for macrophyte dry weight and invertebrate densities obtained from replicate 237-239 samples during the quantitative sampling program and ·the pilot survey (densities expressed as numbers/sample and numbers/g dry wt of macrophyte) . (Appendices 4.1 - 4.6) iv

Appendices - Cont'd Page

5 Adult insects collected in hand-nets and light traps from the shore of Lake Grasmere, 240-246 (Appendices 5.1 - 5.7) 6 Percentage composition of the faecal material in terms of major food categories, and generic composition of the diatom category, for different 247-256 size groups or instars of seven invertebrate species for each month of collection and overall. (Appendices 6.1 - 6.10) v

LIST OF TABLES Table Page

3.1 Factors affecting quantitative sampling in littoral macrophyte beds. 17 3.2 Number of replicate samples required to estimate densities of various invertebrate taxa (+ 20% of the mean) on various macrophytes. 20 3.3 Number of replicate samples required to estimate reliably the dry weight of macrophyte for nine habitat types, and the index of precision when two or three replicates are taken. 20 3.4 Sample type and numbers of replicates to be collected during the quantitative sampling program. 21 3.5 Physical features of the sampling sites. 23 3.6 Sampling dates, sites and numbers of samples collected during the quantitative sampling program. 23 3.7 Results of tests to determine whether numbers/sample or numbers/g dry wt of macrophyte gave significantly less variable density estimates during the quantitative sampling program. 27 3.8 Aquatic macro invertebrates collected from Lake Grasmere, Cass, April 1976 - December 1978. 29 3.9 Percentage composition and mean numbers per sample of macroinvertebrates collected from 13 sites in Lake Grasmere, September 1976 - October 1977. 33 3.10 Percentage representation of species groups at site groups, and the percentage of total invertebrate numbers per sample occurring at each site group. 53 3.11 Percentage occurrence of the taxa of species group 6 among the four site groups. 56 3.12 Species composition of Cladocera in quantitative samples from each site group and total numbers of each species collected, April - October 1977. 57 3.13 Percentage representation of each cladoceran species between the four site groups and overall mean numbers per sample at each site group. 57 3.14 Species composition of Chironomidae in quantitative samples from each site group and all sites combined. 58 3.15 Percentage representation of each chironomid in each of four site groups and overall mean numbers per sample at each site group. 59 3.16 Numbers of samples collected from the site groups, September 1976 - October 1977. 67 3.17 Percentage contribution of the taxa of individual species groups to invertebrate communities of the site groups. 79 3.18 Percentage occurrence of the taxa of species group 6 at each site group. 80 3.19 Contribution of major taxa to invertebrate communities at each site group and for all site groups combined. 81 vi

List of Tables - Cont'd

3.20 Percentage composition of Mollusca at each site group. 81 3.21 Biological index values for major taxa at each site group. 82 3.22 Overall species diversity, species evenness, species richness, numbers of taxa, mean invertebrate densities and numbers of for each site group. 83 3.23 Average values of CC and PSc between sampling dates and for replicate from individual sites for each site group. 85 4.1 Size classes of invertebrates used in faecal analyses. 90 4.2 Standard areas of diatoms used to calculate projected areas of diatoms in faecal analyses. 93 4.3 Composition of periphyton communities on successive lengths of Elodea canadensis stern collected from the eastern sampling area of Lake Grasmere (8 April 1977). 98 4.4 Percentage of Hudsonema amabilis larvae in instars 2 - 5 whose faeces contained various preY,items. 102 4.5 Percentage of H. amabilis larvae in instars 2 5 whose faeces contained each of the major food categories. 104 4.6 Percentage of Triplectides cephalotes larvae in instars 2 - 5 whose faeces contained various prey items. 105 4.7 Percentage of T. cephalotes larvae in instars 2 - 5 whose faeces contained each of the major food categories. 106 4.8 Percentage of Nymphula nitens larvae in three size classes whose faeces contained each of the major food categories. 109 4.9 Percentage of Xanthocnemis zealandica larvae in three size classes whose faecal pellets contained each of the major food categories. 112 4.10 Percentage of X. zealandica larvae in three size classes whose faecal pellets contained various prey items. 113 4.11 Seasonal occurrence of prey items in the faecal pellets of X. zealandica. 116 6.1 Maximum lengths and widths of respiratory trumpets and caudal swim fins of Gressittius antarcticus, Macropelopia languidus, and M. umbrosa. 176

7.1 Taxonomic comparisons of macrophyte-associated invertebrate faunas of Lake Grasmere, New Zealand and other lakes in New Zealand and overseas. 182 7.2 Dominant invertebrate groups in macrophyte zones of some northern hemisphere lakes. 184 7.3 Feeding types (% by number) in relation to habitat in some New Zealand lakes. 187 vii

LIST OF PLATES

Plate Facing page

2.1 Lake Grasmere from the southern end. 6 3.1 The cylinder-sampler. 14

LIST OF FIGURES

Figure Page

2.1 Study area map. 4 2.2 Map of Lake Grasmere. 6 2.3 Lake temperatures and Secchi disc readings. 7 2.4 Seasonal changes in biomass of four species of aquatic macrophyte. 9 3.1 Venn diagram of invertebrate taxonomic representation on four species of aquatic macrophyte. 40 3.2 Invertebrate density and community composition (% by major taxa) on three species of aquatic macrophyte. 41 3.3 Venn diagram of invertebrate taxonomic representation at four depths (1m, 2m, 3m and4m). 43 3.4 Total invertebrate density, community composition (%) and changes in abundance of major taxa between 1 m and 4 m in the Elodea zone. 45 3.5 site group dendrogram. 47 3.6 Species group dendrogram. 49 3.7 Seasonal changes in species diversity, species richness, number of taxa, and species evenness at each site group. 66 3.8 CC and PSc values calculated between adjacent sampling dates for each site group. 68 3.9 Seasonal changes in abundance of major taxa at each site group. 70 3.10 Seasonal changes in total invertebrate density and community composition (%) at site group A. 71 3.11 Seasonal changes in total invertebrate density and community composition (%) at site group B. 73 3.12 Seasonal changes in total invertebrate density and community composition (%) at site group C. 75 3.13 Seasonal changes in total invertebrate density and community composition (%) at site group D. 77 4.1 Skeletal remains used to identify prey items in faecal analyses. 94 4.2 Composition of faecal material of seven invertebrate species from Lake Grasmere in terms of major food categories. 96 viii

List of Figures - Cont'd Page

4.3 Food habits of the last four instars of Hudsonema amabilis. 101 4.4 Sizes of Potamopyrgus antipodarum ingested by H. amabilis. 103 4.5 Food habits of the last four ins tars of Triplectides cephalotes 105 4.6 Food habits of three size classes of Nymphula nitens. 109 4.7 Food habits of three size classes of xanthocnemis zealandica. 113 4.8 Sizes of P. antipodarum found in faecal pellets of x. zealandica. 115 5.1 Hind tibia length versus head width for H. amabilis. 123 5.2 Hind tibia length versus head width for T. cephalotes. 123 5.3 Size frequency distribution' of H. amabilis larvae. 124 5.4 Life-history of H. amabilis. 125 5.5 Size frequency distribution of T. cephalotes. 126 5.6 Life-history of T. cephalotes. 127 5.7 Size frequency distribution of pycnocentrodes aureola. 128 5.8 Life-history of P. aureola. 128 5.9 Life-history of oecetis unicolor. 129 5.10 Size frequency distribution of N. nitens. 131 5.ll Seasonal size frequency distribution of N. nitens. 132 5.12 Life-history of N. nitens. 133 5.13 Life-history of X. zealandica. 134 6.1 Larval prosternal plates and cases of New Zealand Hydroptilidae. 138 6.2 New Zealand distributions of Oxyethira albiceps and Paroxyethira hendersoni. 141 6.3 New Zealand distributions of P. tillyardi, P. eatoni, P. hintoni and P. kimminsi. 142 6.4 Head, gular region, mandibles, antenna and claw of hind limb of P. tillyardi and a claw of the hind limb of P. hendersoni. 145 6.5 New Zealand Hydroptilidae larval and pupal cases. 146 6.6 New Zealand Chironomidae larvae. 158 6.7 New Zealand Chironomidae larvae. 159 6.8 Morphology of adult male Chironomidae. 160 6.9 Male hypopygium of an undescribed Paucispinigera Spa 171 6.10 Wing, antennal tip and hypopygium of male Eukiefferiella sp. 172 6.11 Morphological features of some Chironomidae larvae and pupae from Lake Grasmere. 174 7.1 Food web for invertebrates in the macrophyte zones of Lake Grasmere. 190 ix

ABSTRACT

Ecological studies were made on macrophyte-associated invertebrates at 13 sites within the macrophyte zone of Lake Grasmere, Cass, Canterbury, South Island, New Zealand (43°05'S, l7lo45'Ei 583 m a.s.l.). A pilot survey (14 April 1976) and a quantitative sampling program (September 1976 - October 1977) were carried out using a new cylinder-sampler (which is described). Non-quantitative samples were taken also to provide additional information.

About 113 species of aquatic invertebrates were collected from Lake Grasmere and its immediate environs including at least 75 species that were associated with submerged macrophytes. Chironomidae (17 spp.), Crustacea (16 spp.), and Trichoptera (12 spp.) were best represented. Thirty species were new records for the Cass district.

Cluster analysis was used to simplify the analyses of data from the quantitative sampling program and dendrograms of site groups and species groups were derived. Four site groups were delimited on the basis of similarity in the composition of associated invertebrate communities. Ten species groups were identified on the basis of each species' distribution. The characteristics of the site groups and species groups are discussed. The 13 taxa of species group 6 occurred at all sites sampled and comprised over 95% (by numbers) of the fauna at each site group. Overall, Mollusca (mainly potamopyrgus antipodarum) (55.9%), Crustacea (16.4%) and Coelenterata (Chlorohydra viridissima) (14.8%) were the numerically dominant groups. on macrophytes. Seasonal changes in invertebrate communities at the site groups were considered using diversity indices and community similarity indices.

The trophic relationships of Paroxyethira hendersoni, P. tillyardi, Hudsonema amabilis, Triplectides cephalotes (Trichoptera), Nymphula nitens (Lepidoptera), Xanthocnemis zealandica (Odonata) and P. antipodarum (Mollusca) were investigated using faecal analysis. Emphasis was placed on size- or instar-related dietary changes. N. nitens and T. cephalotes fed primarily on tissue of aquatic macrophytes; the two paroxyethira spp. and P. antipodarum were herbivorous browsers on periphytonj X. zealandica was a predator (especially on Oligochaeta, Cladocera, Acarina and P. antipodarum); and H. amabilis was omnivorous. Notes on the gut contents of Chironomidae are presented. x

Li information was collected for seven species of Trichoptera, N. nitens and x. zealandica. Most species had slow seasonal life-histories, with emergence during spring and summer. Exceptions were Oxyethira albiceps and P. hendersoni (non-seasonal, perhaps bivoltine) and x. zealandica (non-seasonal, predominantly a three year life-history).

Taxonomic studies included the production of keys to larvae of Hydroptilidae and Chironomidae and adult males of the New Zealand Chironomidae. The fifth ins tar larva of Paroxyethira tillyardi is described and the habitats and distributions of the New Zealand of Hydroptilidae are discussed. The male of Eukiefferiella sp. (Chironomidae: Orthocladiinae) is described; illustrated notes on chironomid larvae and pupae from Lake Grasmere are presented and problems in chironomid systematics are discussed. 1

CHAPTER

GENERAL INTRODUCTION

The ecology of invertebrates living in macrophyte zones of lakes in New Zealand has received little attention from biologists, partly because sampling is difficult. Accurate quantitative work is hard to achieve because of the variable density of vegetation, although qualitative data can be obtained simply by sweeping through vegetation with a hand-net (Winterbourn & Lewis 1975). Surveys of the faunas associated with submerged aquatic macrophytes in some New Zealand lakes have been made (e.g., Armstrong 1935, Cunningham et ai. 1953, Winter 1964, Fish 1966, Stout 1969b, 1970, 1975a, 1977, Greig 1973) but only a little is known about the biology of many of the invertebrates concerned (see Marples 1962, Pendergrast & Cowley 1966, Stout 1975a, 1977, winterbourn & Lewis 1975, Chapman & Lewis 1976, Cowley 1978, Winterbourn & Gregson in press). On the other hand, there have been several detailed studies of Hemiptera, Odonata, Trichoptera and Mollusca (e.g., Babington 1967, Winterbourn 1970b, Young 1970, Crumpton 1979, Deacon 1979) •

The present study was initiated in an attempt to overcome this deficiency in the knowledge of the macroinvertebrate fauna and was aimed initially at documenting the taxonomic composition of macrophyte­ associated invertebrate communities to provide a background for further . work on trophic interrelationships and life-histories. Field work was carried out at Lake Grasmere in inland Canterbury, South Island, New Zealand. This lake is readily accessible and is close to the University of Canterbury's Field Station at Cass, but was chosen primarily because it is perhaps the most intensively studied high­ country lake in the Cass district. The scientific endeavour at Cass has been summarised by Burrows (1977) (see also Chapter II of this study) . Aspects of Lake Grasmere that have been investigated include the catchment (McLeod & McLeod 1977, Gibson 1978), physical and chemical features (stout 1969a, b, 1975a, b, 1972, 1977, Irwin 1972), bacterial populations (Ramsay 1972, 1974, 1976), phytoplankton (Stout

1969a, 1972, 1975a, 1977, Flint 1975) 1 zooplankton (Stout 1969a, 1972, 1975a, 1977), periphyton (Greig 1976) and benthic invertebrates (Jamin 1976, Timms in prep.). Invertebrate communities associated with 2 macrophytes have been discussed only briefly by Stout (1969b, 1975a, 1977) and Winterbourn & Lewis (1975). Greig (1976) investigated the production and trophic relationships of a littoral population of a mayfly Deleatidium sp.

The aims of the present study were to examine the invertebrate communities present on macrophytes (native and adventive species), and to elucidate the trophic relationships and life-histories of some of the invertebrates characteristic of the macrophyte zone. An unexpected addition to work originally envisaged was the taxonomic study of larval Hydroptilidae (Trichoptera) and of Chironomidae. The interest in Chironomidae arose when enumeration of light-trap collections of adult insects (initially for Trichoptera) revealed the presence of large numbers of 'unidentified Diptera' which tempted me into trying to identify them. The identification of adults led naturally to attempts to identify larvae and later, because of deficiencies in larval systematics of New Zealand Chironomidae, to more wide-ranging taxonomic investigations (see Stark in press and Chapter VI present study). 3

CHAPTER II

STUDY AREA

2.1 INTRODUCTION

All field work was carried out at Lake Grasmere, Canterbury, near the Southern Alps in the South Island of New Zealand (Figs 2.1 and 2.2) . This lake was chosen because of the relatively extensive background of scientific study available and its proximity to the Cass Field Station, a convenient base.

The purpose of this chap~er is to outline briefly the features of Lake Grasmere, the climate and previous scientific work on the lake.

2.1.1 Location, Formation and Catchment of Lake Grasmere

Lake Grasmere is one of five small lakes situated close together in the catchment of the Waimakariri River (Fig. 2.1) (Stout 1969a). The lake is at a latitude of 43°05' South, longitude 171°45' East and altitude 583 m a.s.l. Like nearby Lake Sarah, it was dammed by morainic ice-eroded rock and postglacial fan-building, and has been in existence since the Poulter glacial advance (about 17 -13 x 10 3 years before present) (Gage 1959, 1977).

The catchment (1,850 hal is primari tussock country and 50% of it is now used for agriculture (Gibson 1978). Agricultural modification of the catchment has occurred during the last 100 year~, although most change has taken place in the last 20 years. In 1956, aerial topdressing and oversowing was begun. In 1964, cultivation and pasture improvement of 75 ha to the west of the lake was undertaken, and in 1973 a border dyke irrigation system began operating so that by March 1977 117 ha were being irrigated. A small reserve of mountain beech forest (Nothofagus solandriivar. cliffortiodes) is present on Long Hill (along the eastern side of the lake). The reader is referred for more detailed information to Gage (1959, 1977), Bradshaw (1977) and Soons (1977)

(geology and geomorphology) i McLeod & Burrows (1977), McLeod & McLeod (1977) and Gibson (1978) (history, farming and catchment of Lake Grasmere). 4

LEGEND

RO.O T[RRACf

er-) F()R[ST I

"-If ,,...lAP I

i-!""'!"'--:o---:---:---....II

Fig. 2.1 The study area. The inset shows the location of Cass (.) and Christchurch (+) in the South Island of New Zealand. Scale of study area map: 1 cm = 1 km. (Modified from Burrows 1977.)

2.2 CLIMATE

Lake Grasmere is located in an area of steep annual rainfall gradient, with high values in the west and low values in the east. It lies on the 1,250 rom mean annual isohyet (Greenland 1977). Monthly variation in rainfall is small, ranging from a mean of 89 rom in February to 130 rom in October at the Cass Field Station (Greenland 1977). Snow may fall in the Cass Basin on a few occasions in most winters, but seldom for any length of time.

High summer temperatures and relatively mild winters (moderated by northwesterly winds) are characteristic of Casso Extremes of shade S

air temperature recorded at the Cass Field Station (1961-1964) were 16°C and 37°C, with a mean annual air temperature for the same period of 9°C.

The climate at Cass and Lake Grasmere is dominated by wind, with a reported average annual wind of 4.9 m sec- 1 (from two years' data at the Cass Field station; Greenland 1977). Observations on wind directions at the Cass Field Station showed that winds from the northwesterly quarter occurred Sl% of the time. Only 20% of the observa- tions indicated calm conditions, and this is likely to be an over-estimate since most observations were taken in the mornings, which are normally calmer than later in the day. Percentage frequencies of winds from other directions were, southeast to northeast 16% and south to southwest 13%.

Greenland (1977) provides further information on the climate of the Cass area.

2 . 3 LAKE GRASMERE

Lake Grasmere (Fig. 2.2 and Plate 2.1) has an area of approximately 63 ha, a maximum depth of lS m (mean depth 7.8 m), and a total length of 1.S km (Irwin 1972, stout 1977). The water level fluctuates by at least a metre during the year, and also differs between years, usually lowest in July/August and highest in spring (Stout 1975a). There is one inlet stream (Ribbonwood Stream), and several inlet springs that originate from neighbouring alluvial fans (stout 1972). The single outlet (Grasmere Stream) leaves the northern end of the lake.

2.3.1 Physical Features

Thermal stratification is a very rare occurrence as the water is frequently mixed by prevailing northwesterly winds. The maximum surface water temperature (recorded in mid lake) has ranged between 19°C and 21°C in different summers, and temperatures in December and January may vary by as much as 4°C between years (stout 1975a). The winter low is usually 34°C and ice may form over all or part of the lake. During the study period, lake water temperatures were within previously recorded limits (Fig. 2.3), and ice covered about seven-eighths and one­ third of the lake's area in the winters of 1976 and 1977 respectively. Ice was thickest along the eastern shore (under the beech forest), and the more wind-exposed and less-shaded southern and western sides of the lake remained ice-free. The lake warmed more quickly and reached a higher maximum temperature in 1977 - 1978 than in 1976 - 1977 (Fig. 2.3). 6.

swampy ground

200 m

Fig. 2.2 Map of Lake Grasmere showing four main sampling areas (N, S, E and W), the 4 m, 8 m and 12 m depth and the point of the lake (15 m) .

Light , measured using a Secchi disc, is variable. Extremes of 0.42 m 1970 following heavy rains) and 8.2 m (May 1970) have been recorded (Stout 1972, 1977), with fluctuations being related to silt loads or plankton densities. The average Sec chi disc value recorded in mid lake by stout (1972) was 2.8 m. Secchi disc readings were taken on occasions during 1976 - 1977 and ranged from 1.7 m (October 1977) to 5.0 m (April 1977) (Fig. 2.3). The average of these values was 3.0 m. Low Secchi disc values during the study period all followed periods of high rainfall (data from weather records kept at the Cass Field Station).

2.3.2 Chemical and Features

The chemical features of Lake Grasmere and its inlets have been discussed by Stout (1969a, 1972, 1975a, 1977), and only the features salient to the present are outlined here.

Plate 2.1 Lake Grasmere from the southern end (opposite).

7

5 u 20 U'l "0

.r:. u 5 ~ (f) 15 0-/\ / 0

o o oI I 10 I

\ ·1976-77

5 / "

o J J A SON D J F M A M J J A S O'N'DIJ 'F'M'A'M' 1976 1977 1978 .

. 2.3 Open-water temperatures and Secchi disc readings taken in Lake Grasmere between July 1976 and May 1978. The dashed line represents temperature measurements of 1976 1977 superimposed on those for 1977 - 1978. Ice was on the lake on 3 July and 25 July 1976 and 13 July 1977.

The frequent mixing and extensive macrophyte beds ensure that the lake water is usually well oxygenated at all depths, although values as low as 36% oxygen saturation have been recorded (following a few days of exceptional summer calm) (Stout 1977) . The pH of the lake water is normally close to neutral, except during algal blooms (which may occur during February or March) when a pH of 8.3 has been recorded (Stout 1969) . Lake Grasmere has amongst the highest values of bicarbonate alkalinity and conductivity for the Cass lakes, but these values are not particularly high compared with values from some North Island lakes. 3 3 Low Ca (6.6 -11. 7 g m- ) and Mg (1.17 - 1.42 g m- ) values, and high Si 3 content (up to 8 g Sio2 m- ) are notable for their likely effects on the decreased occurrence of larger crustaceans and molluscs (Ca and Mg) and increased occurrence of diatoms (Si) (Stout 1977).

Ramsay (1972, 1974, 1976) studied the heterotrophic bacteria in the open water, in water over the macrophytes, on Elodea, and on the mud. 8

She found that the numbers of bacteria in water over Elodea were low, but increased in autumn and winter, and that mature and moribund leaves showed greater bacterial uptake activity than young leaves and new growth. Bacteria tended to increase after zooplankton (but not phytoplankton) blooms and (near an inflowing spring) after rainfall.

The phytoplankton of Lake Grasmere is dominated by diatoms (Asterionella formosa, Diatoma elongatum, Melosira granulata var. angustissima and Cyclotella sp.) with changes in the dominant species from year to year (Stout 1972, 1975a,1977). Filamentous green algae (Mougeotia sp. and Gloeotila pelagica var.) and colonial green algae (Eudorina elegans and Volvox sp.) may occur also, but blue-green algae are not cornmon (Stout 1975a,1977). Values for phytoplankton 3 3 Chlorophyll a as high as 8 mg m- (mean annual value 4.9 mg m- ) have been recorded (Stout 1969a,1972, 1975a, 1977). Lowest values usually occur in May and June (Stout 1972) but marked variations may occur from year to year (Stout 1977).

The zooplankton is dominated by two cladocerans (Ceriodaphnia dubia and Bosmina meridionalis) and several species of rotifer (Stout 1975a). other members of the zooplankton, present at various times of year, are the water mite Piona uncata exigua and young stages of the cyclopoid copepod Eucyclops serrulatus. Seasonal fluctuations of zooplankton, both numbers and species present, can be related, at least in part, to amount and species composition of the phytoplankton (Stout 1975a) .

The adventive Elodea canadensis is the most abundant macrophyte in Lake Grasmere and forms extensive beds in water down to about 7-8 m deep. It is not present in water less than about 0.5 m deep except in occasional small clumps in sheltered areas, and is sparse and stunted at the wind-exposed southern end of the lake. Isoetes alpinus may be found in shallow water, especially along the stony/rocky eastern shore, and also in southern and western areas. The stoneworts, Chara sp. and Nitella sp., grow at depths between 1m and 2 m, especially where the bottom is silty (for example, in places along the eastern and southern shores) . The stoneworts are present in more luxuriant beds below the Elodea. zone in water more than 7 - 8 m deep (Stout 1977). Myriophyllum propinquum and Ranunculus fluitans are present, in a narrow band, along the eastern shore at the upper margin of the Elodea zone, in water about 1 m deep. Patches of M. propinquum may be found in other areas of the lake, especially at the southern end. Potamogeton cheesemanii 9

has a very patchy distribution and was not common during the study period. This species exhibits pronounced die-back in winter (Stout 1975a,1977) and new shoots were observed in silty areas in the northern and western areas of the lake in early November 1976. Seasonal changes in macrophyte biomass were marked (except for M. propinquum) during the present study (Fig. 2.4). E. canadensis had the greatest biomass of the four macrophytes examined and showed the greatest seasonal change. It has perennial shoots that remain green in winter when there was no obvious die-back, although there was some reduction in standing crop (Fig. 2.4). Relatively high water transparency in April 1977 (Fig. 2.3) may have contributed to increased productivity and hence biomass of E. canadensis at this time (Fig. 2.4).

2 (f) ...... (f)N 0 « ...- 0 x ® N ro IE w I- ...... 1 ~ 0... >- 0 I- 0:::: "'0 ~L\.• '0, 0 Q') ,.' \ ..... « , .... _.+ ~ ..+- - -- ...... ",/')!- +- - - "':;.,...- . - , -j( 0

MONTHS

Fig.2.4 Changes in biomass of four species of aquatic macrophyte (Elodea canadensis (E), Isoetes alpinus (I), Ranunculus fluitans (R), and Myriophyllum propinquum (M), (September 1976 October 1977). All standard errors were less than 10 g m- 2 unless indicated. Data from quantitative sampling program (see Chapter III).

Extensive growths of filamentous green algae may be present on macrophytes and stony substrates at certain times of year. During the study period they were most prominent in May and October 1977.

Invertebrate communities associtated with the submerged aquatic macrophytes of Lake Grasmere have been discussed briefly by stout (1969a, 10

1975a,1977) and Winterbourn & Lewis (1975). Work by Jamin (1976) and Timms (in press) has documented the benthic fauna. Greig (1976) investigated the trophic relationships and production of a" littoral population of a mayfly Deleatidium sp. No ecological studies have been made in the lake on invertebrates associated with the macrophytes.

Lake Grasmere contains several species of fish, including brown (Salmo trutta), rainbow trout (S. gairdnerii), long-finned eel (Anguilla dieffenbachii), probably short-finned eel (A. australis schmidtii), upland bullies (Gobiomorphus breviceps) and Koaro (Galaxias brevipinnis) (Stout 1975a).

Lake Grasmere is a Wildlife Refuge for waterfowl and large numbers of birds are often present on the lake (see Stout (1975a) for further details). A number of these birds feed in the lake and nutrients are added to the lake in their faeces (especially by Canada Geese canadensis». 11

CHAPTER III

QUANTITATIVE SAMPLING PROGRAM

3.1 AIMS OF THE QUANTITATIVE SAMPLING PROGRAM

The background to scientific work on Lake Grasmere has been discussed in Chapter II. Various aspects. of the lake, and its associated organisms, have been studied over the years, but information on the invertebrates of the macrophyte zone is limited to preliminary descriptions (Stout 1969a,1975a,1977j Winterbourn & Lewis 1975). Only three studies have been concerned primarily with macroinvertebrates. (1976) studied the ecology of a littoral population of Deleatidium sp. (Ephemeroptera), especially its trophic relationships and predation, but his study was based on wave-exposed stony substrates at the southern end of the lake where a specialised lake fauna exists. Jamin (1976) and Timms (in prep.) have investigated the benthos.

In an attempt to overcome this deficiency in knowledge of the invertebrate fauna, a monthly sampling program was proposed with the following aims in mind:

1. To document the taxonomic of Lake Grasmere.

Preliminary surveys of Lake Grasmere early in 1976 showed that there were many species of invertebrates present in the macrophyte zone that had not previously been recorded. In order to provide the necessary foundation for proposed work on the trophic interrelationships and life histories of the dominant invertebrates it was felt essential that more complete information be collected.

2. To characterise communities on different

The macrophyte zone of Lake Grasmere comprised beds of Elodea canadensis, Myriophyllum propinquum, Isoetes alpinus, Ranunculus fluitans, Potamogeton cheesemani, Chara sp., and Nitella spp. although only the first four named were extensive enough to make regular sampling practical. In particular, a comparison of the communities on the adventive E. canadensis with those on the native M. propinquum was considered potentially interesting. 12

Within the littoral zone of lakes, environmental factors such as water depth, wave exposure, substrate, amount of material and of shading can contribute to variability between communities. It was decided, therefore, that the sampling program should encompass a range of sites within the lake where differences in environmental might be expected. It was hoped that such an approach would yield information of more general applicability to other lakes in New Zealand than would intensive sampling at anyone site.

4. To analyse trends in seasonal changes in invertebrate abundance.

Detailed analysis of the seasonal changes in invertebrate abundance was not considered to be 'a primary aim of this study. The sampling and processing effort required to obtain statistically reliable information would have been prohibitive. Nevertheless, the information obtained, although based on relatively few samples, and low numbers of many species, was thought to be worth presenting with the proviso that the various deficiencies were realised. In addition, it was considered important to gauge the extent that seasonal changes in the invertebrate communities can influence the conclusions drawn from information collected each month when analysed on an annual basis. If seasonal changes are very pronounced the annual overview may not be a good representation of the average situation or it could be influenced markedly by the particular months in which samples were collected. Conversely, where seasonal changes are minimal the effect of dates of sampling on the annual overview should be negligible.

5. To supplement invertebrate life history information.

Most of the invertebrates collected for life history analysis were obtained separately (Chapter V), but the quantitative sampling program was expected to provide additional specimens of the under study.

3.2 METHODS

3.2.1 Sampling Methods (a) The problem of sampling. Sampling of littoral macrofaunal communities has many problems particularly as the littoral macrophyte zone tends to be more diverse than benthic habitats (Welch 1948). When designing a sampling device or assessing the applicability of a given 13

device to a certain situation there are many points to consider. Resh (1979) recently reviewed the sources of error in benthic sampling, sampling reliability, and the role that life history information can play in the development of appropriate designs for benthic investigations. Many of these considerations are applicable also to the sampling of submerged aquatic macrophyte communities. Sampling variability can be dependent upon the choice and operation of the sampling device, environmental features, field and laboratory sorting procedures, and biological features of the study organisms.

A wide range of methods and equipment has been devised for sampling invertebrate communities on aquatic macrophytes (Edmondson & Winberg 1971, Elliott & Tullett 1978, Merritt, Cummins & Resh 1978). For example: i. hand collection (Krecker 1939, Entz 1947, Muller-Liebenau 1956, Matlak 1963, Harrod 1964, Bownik 1970, Laupy 1977);

ii. hand-net sweeps (Stube 1958, Lawton 1970b, Liddle, Happey-Wood & Buse 1979);

iii. mesh bag placed over plant which is then cut off, sealed in bag and removed (Andrews & Hasler 1943, Welch 1948, Rosine 1955, Mrachek 1966, Higler 1980)i

iv. frame, cylinder or box samplers (including grabs and corers) with a variety of closing mechanisms (Whitehead 1935, Macan 1949, Smyly 1952, Gerking 1957, Garnett & Hunt 1965, Gillespie & Brown 1966, Korinkova 1971, Mackie & Qadri 1971, Ladle, Bass & Jenkins 1972, MCCauley 1975, Soszka 1975a, Minto 1977);

v. artificial macrophyte substrates (G1ime & Clemons 1972, Macan & Kitching 1972, 1976, Soszka 1975b, Higler 1977, 1980, Macan 1977).

For quantitative studies, hand collection and hand-net methods are generally considered unsuitable (McCauley 1975). However, there is some evidence that a 'standard sweep' hand-net method can give reasonably quantitative results, although for some invertebrates it can be very selective (because of invertebrate behavioural differences) (Macan 1963, 1966). Several sampling methods are best employed with the aid of SCUBA (notably the mesh bag methods in iii. aboye), otherwise sampling depth is severely limited. Similarly, artificial macrophyte substrates (v. above) are more readily used if SCUBA divers are available, especially for their retrieval. There is some evidence that 14

of phytomacrofaunal community analysis based on artificial substrates can introduce systematic bias. For example, Soszka (1975b) found that Lepidoptera larvae, which burrow in and feed on macrophyte tissue, not unnaturally were absent from plastic 'Potamogeton tus' while present on the real plants. In addition, artificial plants cannot exhibit seasonal changes (in tissue quality, growth form, etc.) which are certain to influence associated animal communities. Furthermore, artificial macrophyte substrates are still relatively untried and, in my opinion, cannot replace sampling of natural macrophyte communities until the various sources of error are recognised and accounted for. Grabs, corers and similar sampling devices (iv) also are not without problems. The equipment required is often more expensive, more complicated and more likely to give trouble. with many samplers, difficulty is in separating the phytomacrofauna from the benthos, there is often a pronounced 'edge effect' and the different growth forms and densities of macrophytes may adversely affect sampling (Resh 1979). However, devices of the grab type seem to offer most promise when it is necessary to obtain samples using a boat and when a range of depths has to be encompassed.

These considerations prompted the design and construction of a cylindrical, pole-mounted sampler that could be operated from a small boat and sample down to a depth of four metres (Stark 1980).

(b) The design and operation of the cylinder-sampler. The sampler (Plate 3.1) is a pole-mounted cylinder (600 mm long x 100 mm internal diameter 'Alkathene' tubing) that can be closed at each end. Stainless steel jaws 2 mm thick (Plate 3.lb), one with a serrated edge, close with a strong rubber strip-powered scissor action across the lower end of the cylinder to cut off and contain the of vegetation and its macrofauna. The top of the comprises a rotating sieve (200 ~m mesh) set at an angle of 45° to the long axis of the cylinder. Four 20 mm-diameter holes in the side wall of the sieve provide through-flow for water as the is lowered and can be closed off by rotation of the sieve. The is attached to an aluminium pole (2.2 m long x 15 mm internal by a hinge at its lower end and a sliding spring-loaded catch near its upper end. A further length of aluminium tubing (2 m long x 15 mm external diameter) when slid 200 mm inside the permanently attached and fastened with a bolt and wing-nut, enables samples to be obtained down to a lake- depth of four metres. (The detachable pole has a brass adaptor that takes the thread of a standard hand-net it also to be Plate 3 . 1 a . General view of the sampler in the set position . Not e : t rigger rope \oJith jaw clamp on lm"er end , rubber loop over end of s i eve rot ation l ever , cyli nder hingE' release catch , a nd rubber " spri ng" arrangement (at left). b . Jaws , in the set pos~tion as seen from below . c . The sieve end with holes open and showing the locating bolt which retains the sieve. d. Sampler in the retrieval position. 15

used to collect nono_oqui:mtitoative samples with a hand-net.) A strong nylon rope, running through brass eye-bolts attached to the aluminium pole, controls sieve rotation, jaw release, triggering and release of 2 the cylinder: from its upper fastening. The sampled area is O. OOB m •

'1'he sampler is set in th,,; following manner (PlatE~ 3.1 a) :

i. The top end of the cylinder is fastened to the pole by the spring-loaded catch.

i1. The :jaws an~ held open by a bx-ass clamp on the end of the triggering rope.

iii. 'I'he sieVe' is rotat.ed so that the four laterally pIa.ced holes are

open, and a rubber loop attached to the trigg~,ring rope is slipped over the end of the lever which controls sieve rotation. It is important th,,1 t the t_riggering rope is positioned as in Plat:es 3. I a and c to facilitate srnoot:h oper'a.tion.

The sequence of operation is as follows:

i. The sampler is allowed to descend over thE~ rnacrophyte bed under

its own wei,,,ht. It was found that ll~ss variable replicate samples were obtodined in this manner, rather than by delicate attoempts, incorporatinq lateral movements, to position the sampler.

it. When the sampler reaches the substrc'lte the Lriqgering rope is pulled and the following three operations occur in sequence -

(al the sieve rotates to close off the four 'irrigation' holes; (b) the jaws are released cutting off and enclosing the sample; and (c) the cylinder is released from its upper fastening.

iii. The sampler is raised with the sieve end at the bottom and the water drain", out: (Plate 3.1 d) .

i v. The sieve. is removE,d by unscrewing a small locatinq bol t and the sample washed into a plastic bag (and preserved if desired). Once the sieve has been replaced the above sequence can be repeated for taking the next sample.

Table 3. 1 summa.x:isE's various considerations (adapt.ed fx-om l<.e5h

1979) involved in tohe quant;it,ative sdompling of animal conununities on mac.rophytes, and methods used in the present study to minimise any inherent or created problems. The sampler incorporates several design featu.t:r;'!s that improve its efficiency. A significant error in

(~dntitative sampling of c macrophytes may arise due to the 'bow 16 wave' of a descending sampler pushing aside the macrophytes or causing the invertebrates with the most developed escape responses to flee. It was found that sampling was easier and a greater number of invertebrates were collected when the sampler was allowed to descend under its own weight and if the sieve was not in place. A lid is essential, however, because once a macrophyte has been cut off it tends to float upward out of the sampler. In order to increase water flow through the sampler (and thus decrease the 'bow wave' and to retain the sample) the four 20 mm-diameter holes were placed laterally on the sieve and the lever system was designed to close these off during sample collection. This design reduced the visible disturbance to the macrophyte bed in the region of the descending sampler and, presumably, led to a reduction of the 'edge effect' (see Elliott 1977). The term effect' describes the unsystematic errors that influence measures of invertebrate density by, for example, the edge of the sampler brushing invertebrates from plant stems or causing those species with the most developed escape responses to flee.

Mesh size of sampling equipment may influence markedly the estimation of popUlation size, distribution and community structure (Malley & Reynolds 1979). The mesh must be fine enough to capture the smallest life history stages required and not so fine that clogging is a problem. The final design of the sieve arrangement on the cylinder sampler was the result of much experimentation. It was found that if the sieve was placed at right angles to the long axis of the cylinder

(i.e. a circular sieve) clogging of the desired ~m mesh with fine sediment and filamentous algae was a major problem when the sampler was used in the manner described above (i.e. the sieve at the bottom when the sampler is retrieved) • A coarser mesh would have allowed too many smaller animals to escape. The solution was to incline the sieve at an angle of 45° to the long axis of the sampler so that the filtering area was increased by 42% and any sediment or algae collected was washed down to the bottom angle of the sieve. This left the greater part of the sieve clear for more efficient drainage.

The sampler was found to be most effective in macrophyte beds where plant growth was vertical. It was difficult to obtain samples where macrophyte stems were very short (i.e. less than about 30 mm, e.g. Isoetes alpinus at some sites) or very long (i.e. greater than about 1 m, e.g. Elodea canadensis in water) or from plant species whose growth forms were not relatively rigid or upright (e.g. Ranunculus 17 fluitans and potamogeton cheesemani). At the times of year when growth of filamentous algae was extensive, difficulty was in obtaining quantitative samples. This was especially so in zones of Elodea canadensis where the filamentous algae and formed an almost impenetrable mat. In such conditions it is very difficult to obtain quantitative samples.

Table 3.1 Factors affecting quantitative sampling of animal communities on littoral aquatic macrophytes (adapted from Resh 1979) and methods used in the present study to overcome the problems.

Factor Problems Remedy vegetation Loss of organisms during Completely enclosed removal; inability to sampler after enclose macrophyte or to powerful close sampler. cutting jaws. Depth Design of sampler limits sampler mounted on extend- working depth. able aluminium with simple triple-action trigger mechanism; sampl­ ing possible down to 4 m lake depth. Disruption of Loss of small organisms and Closeable vents on rotating substrate by those with well-developed sieve to better flow- shockwave of escape responses; incomplete through of water when sampler sampling of macrophyte. sampler descending; care with sampler

Mesh size Coarse miss small life Used 200 ~m mesh angled to history stages. provide efficient Fine - blockage by fine when sampler lifted. sediments leading to reduced sampling efficiency. sampler Large - increased sorting Compromise: the need for dimensions time; may not detect short sorting time biased aggregations (Elliott 1977). the choice in favour of a Small - may not detect small sampler. aggregations; variability increased due to edge effect. Operator Systematic error in popula­ Single operator. inconsistency tion estimates (Needham & Usinger 1956). Number of Time involved vs adequacy Depends on aims of study. replicates of representation of community sampled. weather Can influence micro­ Sample in similar conditions distribution of conditions each time. invertebrates. 18

3.2.2 pilot Survey and Analysis of Sample variabilit~

A pilot survey was undertaken on 14 April 1976 during which 33 samples (Appendix 1) were obtained using the cylinder-sampler. The data from this survey were examined for variability between replicates to determine the number of samples needed to estimate invertebrate abundance to a specified level of precision.

Elliott (1977) and Resh (1979) discussed, in detail, the problems of the size and numbers of samples required to estimate the densities of benthic invertebrate populations to a desired level of precision and these considerations relate equally well to phytomacrofaunal communities. Ideally, the solutions to these problems are dependent upon the aims of the investigation alone, but in practice the choice of sample size and/or the number of replicates is usually dictated by practical, rather than statistical, considerations. A quantitative study aims to estimate the density of each taxon (usually species) in the habitat (commonly used in phytomacrofaunal studies are numbers/m2 of substrate or numbers/g (wet or dry wt) of macrophyte) . On the other hand, a faunal survey usually aims to discover which species are present and perhaps estimate their relative abundance at each site.

The quadrat or sample size most suitable for any particular investigation of the density or distribution of a population is dependent upon the nature of its dispersion; i.e. whether its distribution is random, regular or contagious. contagious or aggregated distributions describe the dispersions of most biological populations (Elliott 1977), which is not surprising considering the spatial variation in the influences of environmental factors.

Several workers (e.g. Beall 1939, Finney 1946, Taylor 1953) have concluded that a small sampling unit is more efficient for estimating population densities than a larger one when the dispersion of a population is contagious. More smaller samples can be collected and processed with the same effort, the catch is more representative since many small units cover a wider range of habitat than a few large units, and many small units (with more degrees of freedom) are more statistically reliable than a few large units. Although a small sample size may be the ideal theoretical solution, there are practical factors which set a lower limit to the dimensions of the sampling unit (e.g. the growth form of the macrophyte species), and, as sample size decreases, the 'edge effect' becomes more pronounced. As samplers become smaller, the edge becomes greater relative to the area sampled 19 and the potential for error in the density estimate increases. The size 2 of the area sampled in this study (0.008 m ) was dictated by the of invertebrates in the habitats to be sampled (estimated from sampling prior to the pilot survey) which influences sample time, and the size of cylinder available for its construction.

Because variation due to contagious distributions is often encountered in sampling natural popUlations, a single small is to be statistically inaccurate. The simplest way to overcome this is to take a large number of replicates (greater than 50). However, this is rarely practicable, especially if samples are to be collected at frequent intervals or from a number of habitats. Elliott (1977) that a standard error equal to ±20% of the mean was reasonable for benthic sampling. Then - standard error 1 the index of precision, 0 0.2 arithmetic mean x where S2 variance, and n = number of If dispersion is contagious (a reasonably safe assumption for most populations) a negative binomial is known to be a suitable model for data obtained from a series of replicate samples. Then -

and n = 1 for 20% error (1) 2 (~+ 1:.) 0 X k where k = common k and f is a measure of excess variance or clumping of - 2) 1 individuals in a popUlation (k = x . As k approaches 0 (k -)0- (0) s2 x the distribution tends to random, and as 1 tends to infinity (k -)0- 0) the k distribution converges to the logarithmic series (~isher, Corbet & Williams 1943). Given the number of (n) actually collected, it is possible to calculate the index of achieved using the following expression:

1 o + k (2) n

Equations (1) and (2) were applied to data (Appendix 1) from replicates collected during the pilot survey (Tables 3.2 and 3.3). The number of replicate samples to estimate reliably the densities of invertebrate taxa within 20% of the mean was very variable, depending upon the habitat , the taxa in question and the units of density used (Table 3.2). When invertebrate density 20

was expressed in terms of numbers per sample, between land 76 replicates were required (depending upon the taxon and the habitat) and when density was expressed as numbers per g dry weight of macrophyte,

Table 3.2 Number of replicate samples (n) required for various invertebrate taxa from different macrophyte habitats to estimate the density* within 20% of the mean. Data from the pilot survey (14 April 1976) (Appendix 1). * n for numbers/sample (n for numbers/g dry wt macrophyte). Habitat codes given in Table 3.4.

Habitat: N E S I E E 2 E I WEI W E 2 W I Number of 8 2 5 4 3 4 2 samples taken: Taxa

Coelenterata 57 (39) 29(38) 15 (13) 41(35) 22(33) 50(43) 6(5)

Annelida 23 (17) - ( - ) 49(78) ( - ) 7 (4) 76 (83) 23 (22) Crustacea 13(22) 26(37) 14(22) 26(37) 23 (24) 27 (25) . 13 (14) Insecta 34(33) 7 (1) 3(2) 17(24) 17 (14) 26(27) 12(13) Acarina 12(6) 1 (10) 3(14) 30(42) 3 (12) 9 (15) 9 (10)

Mollusca 5(4) 18 (5) 5(7) 12(19) 3 (1) 11 (9). 24 (25)

Total 5(4) 14 (3) 5 (6) 13 (21) 4(2) 6(7) 22(23) invertebrates

Table 3.3 Number of replicate samples (n) required to estimate the dry weight in grams of macrophyte within 20% of the mean for nine habitat types, and the index of precision (D) achieved if two * or three ** replicates are taken. Data from the lot survey on 14 April 1976 (Appendix 1). Habitat codes given in Table 3.4.

Habitat: N E S E S I E E 2 E R E I WEI W E 2 W I n 3.3 6.6 6.1 1.3 2.4 2.5 2.0 2.3 <1.0

D* n=2 0.26 0.36 0.35 0.11 0.21 0.22 0.20 0.22 0.02

D**n=3 0.21 0.30 0.29 0.08 0.18 0.18 0.16 0.15 0.02

.~------.--~-- 21

up to 83 replicates were found to be required. Less variability was seen in the total invertebrate density, especially from E. canadensis zones where fewer than eight samples were required to obtain the desired precision. Generally, the number of samples required to estimate reliably the invertebrate density, either as numbers per sample or numbers per g dry weight of macrophyte, within desirable limits of statistical precision, was unacceptably high for the sole investigator since a prime aim was the investigation of several habitat types. Consequently, it was decided to collect between one and three samples from each of 13 habitats (Table 3.4) during the main quantitative sampling program.

Table 3.4 Sample types and number of replicates to be collected during the quantitative sampling program (see Fig. 2.2 for sampling areas and Table 3.5 for details of sampling sites) •

Site Sampling Macrophyte Lake depth No. of samples/ code area (m) month

E E 1 EAST Elodea canadensis 1 1 E E 2 " Elodea canadensis 2 3 E I " Isoetes alpinus 0-1 2 E R II Ranunculus fluitans 1-2 1 E M " Myriophyllum propinquum 1-2 2 WE 1. WEST Elodea canadensis 1 1 W E 2 " Elodea canadensis 2 3 W E 3 " Elodea canadensis 3 1 W E 4 " Elodea canadensis 4 1 W I " Isoetes alpinus 0-1 1 S I SOUTH Isoetes alpinus 0-1 2 S M " Myriophyllum propinquum 0-1 2 N E 2 NORTH Elodea canadensis 2 3

23

Table 3.3 shows the number of replicates required to estimate the dry weight of macrophyte, within 20% of the true mean value, for nine habitats from the pilot survey. Also shown is the influence of the number of replicates (viz., two or three) on the index of precision (D) of the macrophyte dry weight estimate. In most cases (six out of nine) , 22

triplicate (n = 3) samples yielded dry weight estimates within 20% of the true values. Duplicate (n = 2) samples, on the other hand, reduced the precision by an average of just over 4.2%.

3.2.3 The Main

(a) Field procedure. The monthly sampling regime used (Table 3.4) resulted from a consideration of sample collection logistics, the available time per month for sample processing, the variability in samples, and the aims of the study. Given the range of communities to be sampled, the number of replicates of each sample type was severely limited for practical reasons and, as outlined above (p. 21), the number of samples required to obtain a reasonable level of statistical precision (± 20% of the mean density) was far in excess of the number I was able to process in the time available. However, as already stated (p. 12), I considered it more important to investigate a wide range of invertebrate/macrophyte community types than concentrate on one specific community.

The monthly quantitative sampling program was begun in September 1976 and continued until October 1977. Thirteen sampling sites, in four areas of Lake Grasmere (Fig. 2.2) were selected to cover a range of macrophytes and environmental conditions (Table 3.5). Sampling effort was concentrated on Elodea canadensis communities since this macrophyte represented about 90% of the macrophyte-covered area of the lake. The endemic Isoetes alpinus, the most abundant macrophyte in water less than 1 m deep, and, especially, Myriophyllum propinquum (also an endemic species) were expected to provide interesting comparisons with the_adventive E. canadensis. The wide range of habitats investigated was intended to provide data that could be used in comparison with macrophyte communities of other high-country lakes.

A total of 180 samples were obtained with the cylinder-sampler, from the 13 macrophyte habitats in Lake Grasmere on ten occasions during the sampling period (Table 3.6). The samples were collected between 1100 hand 1600 h N.Z.S.T. from four main sampling areas, which were termed North, South, East, and West (Fig. 2.2). Once samples had been collected, they were transferred to 400 x 300 mm plastic bags and preserved with 4% formalin.

Deviations from the proposed sampling regime were due to ice preventing sampling (13 July 1977) or particularly strong north­ westerly winds and excessive growths of filamentous green algae 23

Table 3.5 Physical features of the sampling sites (see Table 3.4 for details of site codes, macrophytes and lake depths and Fig. 2.2 for location of the four main sampling areas). site code Wave action Substrate Remarks

EEl weak-moderate Mostly inorganic but Extensive shading by organic inputs from beech canopy and E E 2 " II overhanging beech hillside (especially E I n If forest canopy. Shore in winter). Myrio­ steeply shelving below phyllum and Ranunculus E M n II 0.5 m. Sand - larg~ comprise a narrow, E R .. 11 rocks and boulders. mixed band near upper margin of Elodea bed. N E 2 moderate Highly organic Extensive macrophyte (decomposition products growth acts as source W E 2 weak of macrophytes). Gentle of, and trap for, W E 3 " slope near shore, organic detritus. several sudden changes Isoetes present in W E 4 " in region of 2 - 4 m. small clumps or in 2 W E 1 moderate Firm to loosely-packed, areas up to 20 m • fine particulate W I " sediments.

S I strong Much organic material strong wave action and deposited by wave possible influence of S M " action (due to prevail­ inlet stream (spring ing NW winds). Uneven, A, Stout 1972) produce stony sediments with stream-like habitat at coarse particulate lake edge. organic material trapped in macrophyte beds.

Table 3.6 Sampling dates and numbers of samples collected from the 13 macrophyte habitats in Lake Grasmere during the quantitative sampling program (see Table 3.4 for key to site codes) .

Site: NE2 SI SM EEl EE2 EI EM ER WEI WE2 WE3 WE4 WI Date

2/ 9/76 3 1 1 1 3 2 1 1 3 1 1 1 2/11/76 3 2 1 1 3 2 2 1 1 3 1 1 2 2/12/76 3 1 1 1 3 2 2 2 1 3 1 1 1 20/ 1/77 3 1 1 3 1 1 3 1 1 1 2/ 3/77 3 2 1 1 3 2 1 1 1 3 1 1 1 8/ 4/77 3 2 1 1 3 1 1 1 1 3 1 1 10/ 5/77 3 1 1 3 1 1 1 3 1 1 1 20/ 6/77 3 1 1 3 2 1 1 3 1 1 1 13/ 7/77 3 2 1 1 3 1 1 1 3/10/77 3 4 1 1 3

30 12 9 6 28 13 10 6 9 30 9 9 9 24

(3 October 1977). The departure from sampling at monthly intervals was primarily due to adverse weather conditions delaying sampling trips.

Samples of R. fluitans (ER, Tables 3.4 and 3.6) were very difficult to obtain due to both the growth form and relative rarity of this plant, which was confined to a narrow band in the eastern sampling area (Fig. 2.2). Sampling of this habitat was discontinued after April 1977.

Lake temperature measurements, Secchi Disc readings and weather observations were made at each time of sampling as variations in these factors may affect sampling efficiency and/or distributions and life histories of many aquatic invertebrates.

(b) Laboratory procedures. Samples were washed through a series of Endecott sieves (3350, 1400, 500 and 200 ~m mesh) to separate animals from plant material, and macrophyte stems were examined to ensure that all invertebrates had been removed. Invertebrates were stored in 70% alcohol to await further sorting. Dry weights (after three days at 66°C) of macrophyte per sample were determined.

Macroinvertebrates were sorted under a stereoscopic dissecting microscope at magnifications ranging from 12.5 to 40 X. Identification of some species, notable Oligochaeta and Chironomidae, necessitated the preparation of slide-mounted specimens. The mounting medium used was lactophenol-PVA incorporating Lignin Pink, which stains chitinised material. Sometimes 10 - 20 seconds in hot KOH was required to clear midge larvae, although animals mounted directly in lactophenol-PVA became clearer with time. Slides were dried for about one week at 66°C, examined, and stored in cardboard slide trays. Macroinvertebrates, once sorted, counted (and measured for use in life-history analyses where applicable: Chapter V) were stored in ethylene glycol.

3.3 RESULTS AND DISCUSSION

3.3.1 Sampling Variability

The pilot survey was the first occasion on which the cylinder­ sampler was used and I found that some expertise was required to obtain good samples. Subsequent modifications to the sampler (including redesigning the sieve end of the cylinder) and practice in its operation, improved the repeatability of sampling. It was probable, therefore, that replicate samples taken in the pilot survey were more 25

variable than similar sets of samples collected later during the main quantitative sampling program, I decided that the best way to test this was to make a paired comparison of indices of precision (D) derived from the pilot survey (column one in Appendices 4,1 to 4.5) with those derived from samples collected on 8 April 1977 during the main quantitative sampling program (Appendices 4.1 to 4.5). As these collections were made at almost exactly the same time of year, when environmental conditions (e.g., macrophyte growth form, presence/absence of filamentous algae) were most likely to be similar, any change in the variability of replicates between dates should have been due primarily to sampling experience.

The statistical test chosen to assess the significance of these comparisons was Wilcoxon's Signed Ranks Test (1945) as described in Langley (1968). The purpose of this simple test is to compare two .. random samples of matched measurements. If there is no significant difference between two sets of paired measurements, chance differences should consist of about equal numbers of plus and minus differences. The test takes into account not only the direction of the differences, but also the size of the differences between matched pairs (cf. Wilcoxon's Sum of Ranks Test in Langley 1968). This feature increases the sensitivity of the test to a point where it compares very favourably with the t Test. Wilcoxon's Signed Ranks Test was preferred to the t Test because it is 'distribution free' and applies to matched measurements (see Langley 1968).

Replicate samples were taken from only four habitats on 8 April 1977 during the quantitative sampling program (Appendices 4.1, 4.2, 4.3, 4.5) . Indices of precision of invertebrate density (for the taxa shown in Table 3.2) from these replicates were compared with those from the pilot survey. In addition, as the accuracy of density measurements is influenced by the units of density, calculations were performed with data

where densities were expressed in terms of numbers/sample (= numbers/ 2 0.008 m ) and in terms of numbers/g dry wt of macrophyte.

In two habitats (N E 2 m and W E 2 m)the probability of there being no significant difference between the precision of the invertebrate density estimates (i.e., numbers/sample and numbers/g dry wt of macrophyte) from the two dates was 5%. This difference was deemed to be probably

significant (P ~ 5%) and inspection of the original data (Appendices 4.1 to 4.5) showed that the quantitative sampling program revealed the lesser sampling variability. The quantitative sampling program gave density

estimates in terms of numbers/sample for habitat S I which were probably 26 significantly better (5% probability of no significant difference) than those from the pilot survey whereas with density expressed as numbers/g dry wt of macrophyte the difference was not significant. For habitat

E E 2 m, there was no significant difference in sampling variability between the two dates for either units of density. However, in no cases did the pilot survey yield indices of precision of invertebrate density estimates that were better than those obtained from like replicates collected on 8 April 1977. Thus, the expectation that sampling performance would be improved by experience probably had some statistical basis.

In order to assess the precision of the invertebrate density and macrophyte dry weight estimates, equations (1) and (2} (p.19) were applied also to replicate samples collected during the main quantitative sampling program. However, it is not possible to apply these statistics to habitats or times when only one sample was collected (e.g., SI, 2 September 1976, Appendix 2.2), Appendices 4.1 to 4.6 show indices of precision of invertebrate density and macrophyte dry weight estimates for replicate samples from eight macrophyte habitat types based on data collected during the quantitative sampling program. Wilcoxon's Signed Ranks Test was applied to paired D values in three ways to determine -

i. whether, for a given month, numbers/sample or numbers/ g dry wt of macrophyte gave a significantly better estimate of invertebrate density for each taxa within each habitat,

ii. whether, for each taxon and total invertebrates, numbers/ sample or numbers/g dry wt of macrophyte gave a significantly better estimate of invertebrate density (all habitats combined), and

iii. whether numbers/sample or numbers/g dry wt of macrophyte gave a significantly better estimate of invertebrate density when data from all habitats, times and taxa were combined.

Table 3.7 presents the results of these significance tests for five habitats. All non-significant results have been omitted and there were insufficient data pairs for habitats E Rand W I to permit the application of Wilcoxon's Signed Ranks Test. Where there was a significant difference between the reliability of invertebrate density estimates on each sampling date, it was usually numbers/sample which gave the better estimate (column i, Table 3.7). The exception was 27

Table 3.7 Significance levels for wilcoxon's signed Ranks Test applied to paired indices of precision (D) of invertebrate density (numbers/sample vs numbers/g dry wt of macrophyte). In each case the unit of density that

gave the significantly less variable (= better) density estimate is given (i) for each sampling date, (ii) for each taxon, and (iii) overall. All non-significant results are omitted. (See Table 3.4 for site codes.)

(i) (ii) (iii) Significance level: 5%* 5%* 1%** Habitat

N E 2 Mollusca, No./g

S I 2/3/77 No./sample No./sample

E E 2 2/12/76, 20/1/77, 2/3/77, 3/10/77 Insecta, No./sample No./sample No./sample

E I 2/3/77, No./g

W E 2 2/3/77, 20/6/77 No./sample

* 5% level difference probably significant. ** 1% level difference almost certainly significant.

habitat EI on 2 March 1977 when numbers/g dry wt of macrophyte gave the less variable estimate. The densities of only two major taxa

(column ii, Table 3.7) were probably significantly better (P ~ 5%) estimated in terms of one unit of density relative to the other. In habitat NE 2 m, the density of Mollusca was probably significantly

(P ~ 5%) more accurately estimated in terms of numbers/g dry wt of macrophyte rather than numbers/sample. This is possibly because Mollusca (here represented exclusively by Potamopyrgus antipodarum and Gyraulus corinna) graze on periphyton on plant surfaces and the available plant surface (which, for anyone plant species, is proportional to the biomass of the plant) is likely to be an important factor in their dispersion. The density of Insecta from habitat

EE 2 m, on the other hand, was probably significantly (P ~ 5%) more 28

accurately assessed in terms of numbers/sample, suggesting that their dispersion in this habitat was on an areal basis (i.e., unit area of lake bottom) and not intimately related to the available area of plant surfaces. Overall data from only two sites (8 I and E E 2 m, column iii, Table 3.7) showed a difference in the reliability of density estimates which was almost certainly (P < 1%) in favour of numbers/sample. Both these sites were relatively difficult to sample, the former due to wave action and resultant patchiness of macrophytes and uneveness of substrate and the latter due to the steeply shelving, rocky substrate.

In conclusion, given that invertebrate density estimates within 20% of the true mean values are usually considered desirable for quantitative sampling programs of this nature; 1. The pilot survey indicated that the numbers of samples required to achieve this precision varied according to the invertebrate taxa, sampling date, macrophyte habitat, and units of density and was considered beyond the limits of practicality for this study.

2. A comparison of samples from April 1976 (pilot survey) and April 1977 showed that sampling variability was reduced in the main quantitative sampling program relative to the pilot survey, suggesting that experience in sampling and modifications to the sieve end of the sampler were factors influencing the reliability of invertebrate density estimates obtained.

3. For most sampling dates, habitats, and taxa in the main quantitative sampling program there was no significant difference between the reliability of density estimates derived from densities expressed as numbers/sample or as numbers/g dry wt of macrophyte. Numbers/sample provided less variable invertebrate density estimates in most of those instances where there was a statistically significant difference.

3.3.2 Community Species Composition

A total of almost 96,000 animals belonging to at least 113 species (Table 3.8)were collected from Lake Grasmere and its immediate environs. These included over 64,000 aquatic invertebrates belonging to about 75 species collected during the main quantitative sampling program from 13 aquatic sites (Appendix 2 and Table 3.9). Ten taxa (Simocephalus vetulus, Liodessus plicatus, Rhantus pulverosus, Helodidae, Polyplectropus puerilis, psilochorema nemorale, pycnocentria evecta, Tipulidae, Paradixa sp., and stratiomyidae) were not collected in quantitative samples but 29 were found in hand-net collections from the lake. None of these taxa (with the exception of S. vetulus, a weed dwelling Cladoceran that was not common during the study period) is considered characteristic of macrophyte zones in lakes. The dytiscid beetles (L. plicatus and R. pulverosus) are most common in ponds, although they do occur at low densities in macrophyte zones of lakes (Pendergrast & Cowley 1966). The remaining species are characteristic of benthic sediments or stony substrates and flowing waters. Five species were represented in all the quantitative samples by two or fewer individuals (Placobdella maorica, Procordulia grayi, Antiporus strigosulus, Triplectides obsoleta and Oecetis iti), and a further five by between three and ten individuals (Deleatidium sp., Zelandobius furcillatus, Sigara arguta, Arrenurus sp. and Physastra variabilis). The damselfly larva, Austrolestes colensonis, was not recorded from the lake during the intensive sampling period but has been collected subsequently (29 September 1980), and was recorded (between March 1969 and January 1970) by Crumpton (1977).

Table 3.8 Aquatic macroinvertebrates collected from Lake Grasmere, Cass, April 1976 - December 1978. L = light-trapped adults only and H hand-collected adults only, from lake shore. * = new record for Lake Grasmere (cf. Stout 1975a». ** = new record for the Cass district (cf. checklist in Burrows (1977».

COELENTERATA Hydrozoa Hydridae Chlorohydra viridissima (Pallas, 1766) PLATYHELMINTHES Turbellaria Tricladida Cura pinguis (Weiss, 1909) ECTOPROCTA Phylactolaemata Plumatellidae Plumatella repens (Linnaeus, 1758) NEMATODA Nematoda indet. * Mermithidae indet. * ANNELIDA Oligochaeta Lumbricidae Eiseniella tetraeda (Savigny, 1826) 30

Lumbriculidae Lumbriculus variegatus (Muller, 1774) * Naididae Chaetogaster sp. * ** Tubificidae Aulodrilus pluriseta (Piguet, 1906) * ** Limnodrilus hoffmeisteri ,Claparede, 1862 * ** Hirudinea Glossiphoniidae Glossiphonia multistriata Mason, 1974 Placobdella maorica Benham, 1906 ARTHROPODA Crustacea Cladocera Bosminidae Bosmina meridionalis Sars, 1904 Chydoridae Alona guttata Sars, 1862 Graptoleberis testudinaria, (Fischer, 1848) * Leydigia ?australis Sars, 1885 * ** Chydorus sphaericus (O.F. Muller, 1785) Daphniidae Ceriodaphnia dubia Richard, 1895 Simocephalus vetulus (O.F. Muller, 1776) Macrothricidae Ilyocryptus sordidus (Lievin, 1848) * Neothrix armata Gurney, 1927 * Ostracoda Cypridae Cypridopsis vidua (O.F. Muller, 1776) scottia insularis Chapman, 1963 * ** Herpetocypris pascheri Brehm, 1929 * prionocypris marplesi Chapman, 1963 * ** Cytheridae Gomphocythere duffi (Hornibrook, 1955) * Darwinu1idae Darwinula repoa Chapman, 1963 * Copepoda Cyclopidae Eucyclops serrulatus (Fischer, 1851) Insecta Ephemeroptera Leptophlebiidae Deleatidium sp. Eaton, 1899 Odonata Lestidae Austrolestes colensonis (White, 1846) * Coenagrionidae Xanthocnemis zealandica (McLachlan, 1873) Corduliidae Procordulia grayi (Selys, 1871) p1ecoptera Eustheniidae Stenoperla prasina (Newman, 1845) Austroperlidae Austroperla cyrene (Newman, 1845) Gripopterygidae Zelandobius furcillatus Ti11yard, 1923 * 31

Hemiptera Corixidae Siqara arquta (White, 1878) Diaprepocoris zealandiae Hale, 1924 * Coleoptera Dytiscidae Liodessus plicatus (Sharp, 1882) Antiporus striqosulus (Broun, 1880) * Rhantus pulverosus (Stephens, 1828) * Helodidae indet. Diptera Tipulidae Zelandotipula sp. H Leptotarsus (Macromastix) minutissima (Alexander, 1922) H Limonia (Dicranomyia) otaqensis (Alexander, 1924) ** H L. (D.) ?vicarians (Schiner, 1868) L L. (D.) sp. A ** H L. (D.) sp. C ** H Metalimnophila sp. H Aphrophila neozelandica.' (Edwards, 1923) HL Erioptera (Trimicra) pilipes (Alexander, 1922) HL Amphineurus 4 spp. (none named) ** HL (2 spp.) Molophilus sp. H Chironomidae Tanypodinae Pentaneura harrisi Freeman, 1959 ** H Ablabesmyia mala (Hutton, 1902) * ** Gressittius antarcticus (Hudson, 1892) * Macropelopia lanquidus (Hutton, 1902) L M. umbrosa (Freeman, 1959) * ** Podonominae Parochlus spinosus Brundin, 1966 * HL pupae Diamesinae Maoridiamesa harrisi Pagast, 1947 H Lobodiamesa campbelli Pagast, 1947 H Orthocladiinae Syncricotopus pluriserialis (Freeman, 1959) * Cricotopus zealandicus Freeman, 1959 * Metriocnemis lobifer Freeman, 1959 ** H Corynoneura donovani Forsyth, 1971 * ** Lymnophyes vestitus (Skuse, 1889) * ** Orthocladiinae indet. sp.A * ** Orthocladiinae indet. sp. B * ** Orthocladiinae indet. sp. C * ** Chironominae Chironomus zealandicus Hudson, 1892 Chironomus sp.a * ** Xenochironomus canterburyensis (Freeman, 1959) * ** Cladopelma curtivalva (Kieffer, 1917) * ** Harrisius pallidus Freeman, 1959 H Polypedilum pavidus (Hutton, 1902) ** L P. canum Freeman, 1959 ** H Tanytarsus vespertinus Hutton, 1902 '* Calopsectra funebris (Freeman, 1959) * '** Dixidae Paradixa sp. '* Stratiomyidae stratiomyidae indet. 32

Trichoptera Hydropsychidae Aoteapsyche colonica (McLachlan, 1871) HL A. tepoka (Mosely, 1953) HL Polycentropodidae Polyplectropus puerilis (McLachlan, 1868) '" Rhyacophilidae Hydrobiosis umbripennis McLachlan, 1868 L H. parumbripennis McFarlane, 1951 L H. harpidiosa McFarlane, 1951 L H. frater McLachlan, 1868 L H. clavigera McFarlane, 1951 L psilochorema nemorale McFarlane, 1951 Ps. bidens McFarlane, 1951 L Ps. leptoharpax McFarlane, 1951 L Costachorema xanthoptera McFarlane, 1939 L c. callistum McFarlane, 1939 L Conoesucidae pycnocentria evecta McLachlan, 1868 Pycnocentrodes aureola (McLachlan, 1868) Hydropti lidae Oxyethira albiceps (McLachlan, 1862) Paroxyethira hendersoni Mosely, 1924 P. tillyardi Mosely, 1924 * ** Leptoceridae Oecetis iti McFarlane, 1964 * o . unicolor (McLachlan, 1868) * Hudsonema amabilis (McLachlan, 1868) * Triplectides cephalotes (Walker, 1852) T. obsoleta (McLachlan, 1862) * Lepidoptera pyralidae Nymphula nitens (Butler, 1880) Acari Cryptostigmata Hydrozetidae Hydrozetes lemnae (de Coggi, 1899) * ** Malaconothridae Trimalaconothrus novus (Sellnick, 1929) * ** Astigmata Acaridae Rhizoglyphus robini Claparede, 1869 '" ** Prostigmata Arrenuridae Arrenurus (Arrenurus) sp. * ** Pionidae Piona (Piona) uncata exigua Viets, 1949 Mollusca Gastropoda Planorbidae Gyraulus corinna (Gray, 1850) Physastra variabilis (Gray, 1843) Hydrobiidae potamopyrgus antipodarum (Gray, 1843) Bivalvia Hyriidae Hyridella menziesi (Gray, 1843) Sphaeriidae Sphaerium novaezelandiae Deshayes, 1851 Table 3.9 .M.acroinvertebr&tl!s collected in quantitative samples, Lake Grasmere .. Cass, September 1976 - October 1977. composition by number of total fauna

collected from e.s.ch site and, in brackets f mean numbers per sample (oo O~OO8m:l).. * = <: O.H. of the total fauna or individuals. per sample. - III. not present. Site codes as in Table 3.4, raw collection data in Appendix 4. (Plumatella repens was recorded as present at every site but 'Was not quantified.)

S I T E S

Taxa N E 2 S I S H EEl E E 2 E I E H E R WEI II F 2 WE 3 WE 4 II I

COELENTERATA Chloxohydra viridissim.a 3.1 (1l.1) 2.3 (5.3 0.8 (3.1) 8.1 (42.7) 9.0 (35.0) 2.6 (3.9) 9.4 (16.~) 1.3 (2.0) 1.4 (6.1) 22.9 (80.2) 49.3(300.9) 49.8(267.2) l.0 (4.8) PLATYHELMINTHES . Cura pinguis 0.1 (0.4) 0.3 (0.6 0.1 (0.4) 0.4 (2.0) 0.3 (l.0) 0.5 (0.7) 0.4 (0.6) 0.1 (0.2) 0.1 (0.6) 0.1 (0.4) 0.2 (0.9) 0.1 (0.7) 1.1 (4.9) NEKATODA 0.3 (1.0) l.2 (2.6 0.5 (1.9) 0.1 (0.5) 0.1 (0.4) 0.3 (0.5) 0.1 (0.1) 0.1 (0.2) 0.5 (2.2) 0.1 (0.2) - - -- 1.4 (6.6} OLlGOCHAE'rA (except C/u:lec"9""ster) 5.2 €l8.3) 7.4 (16.8 LO (4.0) 0.9 (4.0) L9 (7.3) 0.4 (0.6) . (0.11 0.3 (0.5) 7.7 (34.9) 6.1 (2l. 5) 0.9 (5.3) 0.5 (2.4) 1.0 (4.7) Cilaetogaster sp. 1.4 (5.0) 2.1 (4.8 1.4 (5.2) 0.2 (1.0) 0.3 (1.0) 0.3 (0.5) 1.7 (2.9) 0.2 10.3) 6.0 (27.3) 2.0 (7.2) 2.2 (13.2) 3.5 (lB.7) 1.5 (6.9) HIRUDINEA Glossiphonia multistriata 0.1 (0.3) ( - - (0.2) ------(0.1) -- -- Placobdella l1l43orica I ) ·- · - ·------· - -- - - CRUSTACEA · · Cladocera B.O (28.5) B.l (lB.3) 4.B (lB.6) 0.6 (3.2) 1.0 (4.0) 5.4 (B.O) 2.8 (4.8) 1.1. (1.7) 7.6 (39.4) 8.6 (30.0) 9.2 156.3) 3.2 (17.1) 5.9 (26.8) Ostracoda ( ) 0.3 (0.6) 0.9 (3.6) 0.8 (4.2) 0.7 (2.6) 0.8U1. 7) 0.1 (0.2) 0.2 (0.3) l.0 (4.3) ( ) (0.3) 0.1 (0.4) 6.2 (28.3) Copepoda · · · · · Eucl}clops serrula tus 1B.4 (65.3) 3.8 (B. i) 3.4 (13.0) 6.2 (32.7) 6.5 (25.5) 23.4(34.8) 5.0 (8.7) 5.2 (8.2) 8.3 (37.0) 9.0 (31.7) 11.4 (69.B) 7.9 (42.1) 12.8 (58.6) INSECTA Ephemeroptera Deleatidium "p. - ( ) -- --- 0.1 (0.1) - (0.1) ------Odonata · · · Procorduli;;. gray! ( ) ------. (0.1) -- x.ant:hocnernis zealandicrl ( · ) ------0.2 (0.8) 0.1 (0.3) (0.2) - - - - Plecoptera · · · Zelandobius Eurcillatus - 0.1 (0.2) (0.1) - -- - 0.2 (0.3) ------HeMiptera · Diaprepocoris zealandiae ( ) 0.1 (0.2) 0.1 (0.2) (0.2) 0.1 (0. ) 0.1 10.1) - - - ( ) ------Sigara arguta · ( · ) - - (0.1) ·-- - -- 0.1 (0.1) 0.1 (0.2) -- ·- ·- (0.2) -- - Coleoptera · · · · Antiporus strigosulus (0.2) (0.1) --- -- ~ -- 0.1 ------Trichoptera · O¥l}ethira albiceps 0.1 (0. ) ------0.2 10.8) ------(0.1) nendersonl 0.9 (3.1) 3.8 (8.6) 3.0 (11. 7) 0.5 12.5) 0.2 (0.8) 2.4 1).6) 3.2 (5.5) 0.2 (0.3) 0.8 13.9) 0.8 12.B) 0.4 (2.2) 0.3 (1. 7) 5.5· (25.0) P. ( ) ( ) - - 0.9 (4.8) 0.6 12.4) 1.1 (1.7) 1.4 (2.5) 3.1 (4.8) 0.1 (0.2) 0.1 (0.2) - - 0.1 (0.3) 0.3 (1. 4) . Pycnocentrodes aureola · · · 0.2 (0.9) 10.2) (0.1) 0.1 10.2) - - - (C .1) ------Triplectides cephalotes ( ) 0.1 (0.3) 0.3 (1.2) ·- - ·-- 0.1 (0.1) - 0.1· Ie ',) ( ) - - - - 0.2 10.8) T. obsoleta · ·-- - [0.1) - - ( . ) - -- - ·-- · ------Hudsonem.:.l amabilis ( I - - 0.1· (0.4) ,(0.2) 0.2· (0.9) 0.3 (0.5) 0.6 (1.0) 0.5 (0.9) (0.1) ------Oecetis unicolor · · 0.1 10.2) - --· ------· (0.1) ------0... iti - - ( ) ------. - - ·------(0.1) Lepidoptera · · · Nymphuld: nitens ( ) 0.3 (1.0) 0.1 10.3) (0.1) 0.1 ( . ) 0.6 ILl) ------0.1 (0.3) Oiptera. · · · Chironomidae O.B (2.B) 5.2 (11.7) 6.0 (23.3) 0.4 (1.9) 0.5 (2.0) 3.3 (4.8) 1.4 (2.5) 1.3 (2.0) 2.5 (11. 4) 1.3 (4.6) 1.2 (7.2) 1.1 (5. B) 3.3 (15.2) Acari Arrenurus n.sp. ------( . 1 -- (0.1) ( ) - - - - (0.1) piona uncata exigua 1.8 [6.3) 0.9 12.1) 0.5 (1.8) 1.7 (9.0) 2.0· (7.7) 1.0 (1.5) 2.0 (3.5) 2.3 (3.7) 0.5· [2.4) 1.9· (6.7)· (5.0) 0.9 (4.7) 0.4· (2.0) (7.3) (9.2) 3.4 :;,;v.v.~. lemnae 0.9 13.11 4.5 (10.2) 2.2 [8.7) 2.3 (12.0) 1.4 (5.5) 4.3 (6.3) 4.7 (8.2l 5.1 (9.0) 0.5 (2.3) 0.8 P.7) 1.2 1.7 U5.3) novus 0.5 (1.9) 0.3 (0.7) 0.9 (3.6) 0.3 (1.5) 0.1 10.3) 3.4 (5.0) 0.3 (0.5) 0.4 (0.7) 0.2 (0.6) 0.2 10.8) 0.1 (0.7) 0.8 (4.3) I 4.0 (4.3) Mollusca corinna 10.7 (39.0) B.l (18.3) 5.7 (22.1) 0.7 D.7) 1.9 (7.3) 1.5 (2.2) O.S (1.3) 1.0 (2.2) 9.3 142.3) 7 •• (26.1) l'"3.7 (22.4) 5.2 127.7) 5.0 (23.0) variabilis ------0.1 (0.2) 0.1 ( • ) ------PotaJrOpyrgus antlpodarum 47.7(169.2) 50.5(113.8) 67.7 (262.7) 75.5(398.3) 72.7(283.0) 48.3(71.9) 16~: ~ (1i~:!: 76.9(189.9) 52.9(240.2) 38.5(134.9) 9.3(117.4) 5.0(134.3) 45.3(207.8) Sph{1erium novaezelandiae - - 0.6 (1.3)1 0 • 1 (0.6) 0.4 (2.3) 0.4 (1.5) 0.1 ( • ) 0.1 (.) i 0.2 (0.7) 0.1 (0.2) 0.1 (0.2) - - 1.4 (6.3)

w w 34

3.3.3 Known Habitat of Some Freshwater Invertebrates

The purpose of this section is to review previous knowledge of the habitat requirements of many of the species collected in quantitative samples from Lake Grasmere, in order to provide a background for interpretation of the association of species in species groups (see p. 52).

PLATYHELMINTHES

Cura pinguis, a triclad, is found under stones, on plants, in ponds with muddy bottoms and in stable rivers (Nurse 1950) and also on dense submerged macrophyte beds in lakes (Winterbourn & Lewis 1975).

ANNELIDA Oligochaeta Lumbriculus variegatus and Eiseniella tetraedra have been recorded from macrophyte zones in other lakes in New Zealand (Winterbourn & Lewis 1975) and the former was the most abundant oligochaete in the mud below the macrophyte zone of Lake Grasmere (Jamin 1976, Timms in prep.). L. variegatus is known from a wide range of habitats but favours quiet reaches of flowing waters and ponds with silt, mud and roots in the littoral zone (Pickavance 1971). A Lumbriculus-dominated community is not considered indicative of organically polluted conditions (Brinkhurst 1965) but rather of waters supporting abundant submerged vegetation and/or a substrate of sand, silt, gravel or coarse detritus (Winterbourn

& Stark 1978). The naidid oligochaete, Chaetogaster sp .. , is represented in the psammon fauna and flora living in interstitial water between grains of sand in the substrate) (Winterbourn & Lewis 1975) but may be found in the mantle cavity of Gyraulus sp. (near punedin especially, according to Marples 1962).

CRUSTACEA Cladocera Bosmina meridionalis is a filter-feeding planktonic cladoceran found in many lakes and reservoirs throughout New Zealand. Graptoleberis testudinaria on the other hand is characteristic of macrophytic habitats where it glides over stems and leaves, grazing on bacteria and small particles on their surfaces (Lewis 1976). Alona guttata is characteristic also of macrophyte zones (stout 1975a, Winterbourn & Lewis 1975) .. Chydorus sphaericus has similar habits to G. testudinaria except that it feeds on filamentous algae and is not infrequently recorded from the open-water plankton (Lewis 1976). Ceriodaphnia dubia is found in most lakes and ponds throughout New 35

Zealand and is regarded as planktonic. Ilyocryptus sordidus is a cosmopolitan macrothricid cladoceran that has a truly benthic habit living in soft bottom deposits of ponds and lakes, and Neothrix armata frequents similar habitats although it has been recorded only in benthic samples from several lakes in the Rotorua (North Island) district (Lewis 1976) prior to this record.

Ostracoda Six species of ostracods were identified from Lake Grasmere. All are characteristic of benthic (Scottia insularis, Herpetocypris pascheri, Prionocypris esi, Gomphocythere duffi and Darwinula repoa) or macrophytic habitats (Cypridopsis vidua and D. repoa) (Chapman 1976) •

Copepoda Eucyclops serrulatus, an herbivorous cyclopoid copepod that creeps over macrophyte stems or benthic substrates, was the only copepod recorded from quantitative samples. It is known from lakes, ponds, rivers, and even brackish waters in coastal situations (Chapman & Lewis 1976) .

INSECTA Ephemeroptera The leptophebiid mayfly Deleatidium sp., probably New Zealand's most common aquatic insect genus, is widespread in streams and rivers and is found also in lakes on wind-exposed shores (Penniket 1969, Winterbourn & Lewis 1975, Greig 1976). The taxonomy of Deleatidium spp. is most uncertain.

Odonata Austrolestes colensonis is common in small pools and swamps and may be found at the of lakes where there are plant species suitable for oviposition. It is not found in habitats where there is significant water movement or wave action (Crumpton 1975, 1977). Xanthocnemis zealandica is found in similar habitats to A. colensonis and is distributed widely in New Zealand. It is associated usually with aquatic vegetation and, unlike A. colensonis, may be found in slow­ flowing streams (Crumpton 1977). Within-habitat distribution of X. zealandica is patchy in heterogeneous environments (e.g., Typha beds, Lake Sarah) but more even in uniform habitats (e.g., Isaac's Pond) (Deacon 1979). The endemic anisopteran, Procordulia grayi, is fairly widespread in bottom sediments of ponds and lakes both in the Canterbury high-country, and in lowland areas of Canterbury and Westland (Crumpton 36

1977). More generally, larvae may be found in suitable habitats in New Zealand south of the Waikato (North Island) in lakes, ponds and occasionally deep rivers. Small larvae frequent deep water (3 m +) whereas later instars occur nearer the shore (R.J. Rowe, pers. comm.).

Plecoptera The only stonefly recorded in quantitative samples, Zelandobius furcillatus, has been recorded from open stable streams and, like Deleatidium, wind-exposed lake shores (Stout 1977).

Hemiptera Sigara arguta is the most common corixid in New Zealand and is found in most still waters, such as in sheltered places in large lakes, along margins of estuaries (sometimes brackish), in ponds and in slack­ water areas of rivers and streams ('except those in the South Island with shingle beds) (Young 1962). In lakes and large ponds it is often associated with Diaprepocoris zealandiae. D. zealandiae is common in high country lakes of the South Island and in larger ponds, lagoons, and canals in coastal areas. It is less common in slow water areas of deep streams. The presence of aquatic macrophytes (especially Elodea, Myriophyllum or Ranunculus) or Typha beds are an essential feature of its habitat and the water must be more than 0.3 m deep, at least in places (Young 1962).

Coleoptera The only dytiscid in quantitative samples, Antiporus strigosulus, has been recorded mostly from ponds, tarns, and among vegetation in lakes (Stout 1969b,1977), temporary rainwater ponds and some streams (Ordish 1966). The genus (as unidentified larvae) has been found also in thermal waters of Lake Rotowhero up to 34°C (Winterbourn 1968) .

Diptera : Chironomidae (see Chapter VI)

Trichoptera (see Chapter VI for Family Hydroptilidae) Cowley (1978) stated that the sericostomatid caddisfly pycnocentrodes aureola was found on rocks in streams, although he did examine some material from the rocky shore of Lake Grasmere. Since the larval case is always made of small stones and sand (and is relatively heavy for the size of the larva), the larva must be associated with substrates including fine mineral particles.

Five species of leptocerid caddisflies were collected from macrophytic habitats of Lake Grasmere but only three were common. Oecetis unicolor was first studied by Babington (1967) in Lake Rotorua 37

(North Island) where it was found in water up to 0.6 m living in the upper 6 - 7 mm of sand on the bottom, and also in loosely-growing macrophytes close to the substrate. It is common in many lakes that have clean sandy areas of beach and it occurs also in rivers and streams in similar habitats (Cowley 1978). M.L. Ling (pers. comm. to Cowley 1978) recorded its presence in the Ohau Channel between Lakes Rotorua and Rotoiti in water up to 2 m deep where there was moderate water flow and Timms (1980) recorded it down to 46 m in Lake Rotoiti (South Island).

Hudsonema amabilis is probably the most widespread of the leptocerid caddis flies found in New Zealand. It is present in rivers and streams and also in lakes especially (but not exclusively) near the shore where there is some wave action, and near inlet or outlet streams (Cowley 1978) but may occur at greater depths (e.g., down to 20 m in Lake Rotoroa, Nelson lakes, South Island (Timms 1980».

One of the largest leptocerids in New Zealand is Triplectides cephalotes, which is one of the most common still-water caddisflies in the country. It is found throughout New Zealand in lakes that have suitable macrophyte zones but is absent from high-altitude tarns. It is known also from temporary ponds and neglected (?) swimming pools (Cowley 1978). Babington (1967) noted that in the macrophyte zone of Lake Rotorua (North Island), larvae were most common in water less than 1 m on plants less than 0.3 m high. However, B.T. Coffey (pers. comm.to Cowley 1978) observed T. cephalotes larvae on Lagarosiphon down to a depth of 6 m in the Nelson lakes (South Island) and Timms (1980) recorded it down to 21 m in Lake Rotoroa and 12 m in Lake Rotoiti (Nelson lakes).

Lepidoptera Nymphula nitens, New Zealand's only moth with an aquatic larva, is almost always associated with macrophytes (Marples 1962) although it has been recorded from benthic habitats (Forsyth 1978, Timms 1980 and in prep.). In these benthic situations, macrophytes must have been nearby because the larva is an pbligate herbivore (feeding on living plant material) (see Chapter IV). This species is widespread in New Zealand and is present also in South Australia (Pendergrast & Cowley 1966) .

ACARI Piona uncata exigua is the most commonly occurring water mite in New Zealand and is usually found in small lakes or large ponds. This 38 species has strong swimming setae on its and, unlike most other

aquatic mites, may be well represented in the plankton (Stout 1969a) 0

Hydrozetes lemnae is an aquatic oribatid cryptostigmatid) mite that inhabits rootlets, stems and leaf sheaths of aquatic plants. The Group Oribatei comprises mostly non-predatory mites that feed only on dead moss or macrophyte tissue (i.e., mycophages or saprophages) in moist to submerged habitats (Krantz and Lindquist 1979). This particular species is widespread in New Zealand (Dr G.W. Ramsay pers. comm.) and one of the few published records is from the outflow channels of hot springs near the Hurunui River (Canterbury, New Zealand) at temperatures up to 41°C (Stark, Fordyce & winterbourn 1976). Another oribatid, Trimalaconothrus novus, has been recorded from localities near Queenstown and Dargaville (Dr G.W. Ramsay pers. comm.) and hot springs near the Hurunui River (Stark, Fordyce & winterbourn 1976). Members of this genus are knO\ffi to feed on green plants and are common on mosses in moist, semi-aquatic or submerged habitats (Krantz & Lindquist 1979). In the hot springs, T. novus was found in association with the blue­ green algal mat up to 41°C. In general however aquatic oribatids in New Zealand, although found in a variety of habitats, have been little studied (Stout 1976) .

MOLLUSCA Gastropoda Gyraulus corinna, the most common planorbid mollusc in New Zealand freshwaters, is an endemic species but little is known about it. G. kahuica, regarded by some .g., Ponder pers. comm. to Winterbourn 1973) as a subspecies of G. corinna, acts as an intermediate host for larval trematodes (Winterbourn 1973).

Physastra variabilis has a wide distribution in still waters in New Zealand but rarely is common. It is particularly sparse in the vicinity of towns and in water bodies that are influenced by man's activities. P. variabilis is most common in lakes and ponds in remote areas, but seems to be unable to compete effectively with the closely related Physa sp. which has replaced it in certain habitats (Winterbourn 1973).

Potamopyrgus antipodarum, a hydrobiid gastropod, is certainly the most common and widespread of New Zealand's freshwater molluscs and is found in almost all kinds of fresh waters except temporary ponds. It also inhabits brackish waters (Winterbourn 1970b, 1973). 39

Bivalvia

Sphaerium novaezelandiae is the most common sphaeriid clam in New Zealand and may be found in mud from both vegetated and unvegetated areas of lakes as well as in streams that have suitable substrates (Winter bourn 1973).

3.3.4 Invertebrate Community Relationships

(1) Invertebrate communities on different macrophytes

The invertebrate communities found on four species of submerged aquatic macrophyte were examined during the quantitative sampling program: the adventives Elodea canadensis (142 samples) and Ranunculus fluitans (10), and the endemics Isoetes alpinus (42) and Myriophyllum propinquum (19) (see Mason 1975). Thirty-seven invertebrate taxa were recognised.

Only 25 invertebrate taxa (Fig. 3.1) were collected in samples of R. fluitans and these taxa (with the exception of Xanthocnemis zealandica (e» were present also on the three other macrophytes.

Thirty-six invertebrate taxa were collected from samples of E. canadensis (Fig. 3.1). The E. canadensis zones comprised about 90% of the macrophyte-covered area of Lake Grasmere and, therefore, provided a very extensive habitat for invertebrates. Oecetis iti (which was found only on Isoetes) was the only species not recorded from E. canadensis and Placobdella maorica and Procordulia grayi were unique to it. These last two species were found only occasionally and their apparent absence from other plants may reflect the fewer samples collected from other macrophytes.

I. alpinus was the habitat of 35 taxa, lacking only the two species unique to E. canadensis (d, Fig. 3.1). However, many of the taxa recorded from this plant were not common on it (groups b, c, d, e and f, Fig. 3.1) (see also Table 3.8).

The fauna collected from M. propinquum (30 taxa) comprised only the 24 taxa shared by all four macrophytes and six shared with Elodea and Isoetes.

The proportion of insects in those groups not present in communities on all macrophytes (b f, Fig. 3.1) was high, suggesting that the insects had quite specific habitat requirements (at least in terms of macrophyte preferences) . 40

(a) Chlorohydra viridissima Cura pinguis Plumatella repens NEMATODA OLIGOCHAETA Chaetogaster sp. Glossiphonia multistriata CLADOCERA OSTRACODA Eucyclops serrulatus Sigara arguta Diaprepocoris zealandiae N=37 Oxyethira albiceps Paroxyethira tillyardi P. hendersoni Hydrozetes lemnae Hudsonema amabilis Trimalaconothrus novus Nymphula nitens Gyraulus corinna CHIRONOMIDAE Potamopyrgus antipodarum Piona uncata exigua Sphaerium novaezelandiae (b) Antiporus strigosulus Triplectides cephalotes Zelandobius furcillatus Oecetis unicolor Pycnocentrodes aureola Physastra variabilis (c) Deleatidium sp. Arrenurus sp. Triplectides obsoleta (d) Placobdella maorica Procordulia grayi (e) Xanthocnemis zealandica (f) Oecetis iti

Fig. 3.1 Venn diagram of the distribution of invertebrate taxa among the four most abundant macrophyte species in Lake Grasmere. Data from quantitative samples (April 1976 - October 1977). Numbers of samplesj E. canadensis = 142, I. alpinus 42, M. propinquum = 19 and R. fluitans 10.

The units of invertebrate density used to compare communities on different macrophytes (or even on the same macrophyte between times) can have a marked effect on comparisons and subsequent interpretation (.Fig. 3.2). When total invertebrate density was expressed as numbers/m2 of lake bottom «a) Fig. 3.2) E. canadensis appeared to represent the most productive habitat, followed by Myriophyllum and Isoetes. Elodea tended to grow in the densest beds; whereas the growth of Myriophyllum was much more open. Isoetes could reach high stem densities but was more often clumped and the inclusion of benthic invertebrates in samples was significant.

When densities were expressed as numbers/g (dry wt) of macrophyte «b) Fig. 3.2) the highest values were found on Isoetes because this macrophyte had a lower biomass per sample than either Elodea or Myriophyllum but invertebrate numbers were not decreased 41

(al (bl (cl (al (bl (cl - 100 3 r- - Coelenterata '"~ ~ 0 x Annelida >- ~. r- ~ (j) 2:: \I) z C 10 E Crustacea CL w E4 ~ - :J 0 CL 2!2 - 0 0 \I) Insecta ~ - Acarina ls> w 0 '0 L- r- ..0 c-- 0 . c: W ~ ~ '0 r- u 5 Mollusca a:: -2 ~ 1- ~ w 0 E 0 > ~

0 0 o ®CD@ MACROPHYTE

Fig. 3.2 Invertebrate density and community composition (%) from samples of Elodea canadensis (E) 143 samples, Isoetes alpinus (I) 42 samples and Myriophyllum propinquum (M) 19 samples collected from Lake Grasmere between April 1976 and October 1977.

proportionately. In addition, since Isoetes stem lengths were short, invertebrates characteristic of benthic habitats probably were included in samples. Elodea and Myriophyllum, which had similar general growth forms, harboured communities of similar density.

The third measure of invertebrate density, numbers per length (metres) of plant stem «c) Fig. 3.2) altered the ranking once again. In this case the order of increasing invertebrate was reflected by available area of plant surface, especially as a substrate for grazers

on 'Aufwuchs ' (= diatoms, etc., on plant surfaces). Myriophyllum, with its finely' divided leaves. (cf. Elodea) presented the greatest surface area, and Isoetes, with erect, spike-like leaves crowded on short stem bases, the least. There was a good relationship between numbers of invertebrates/g (dry wt) of macrophyte and numbers per length of plant stem for macrophytes of similar growth form (I.e., Elodea and Myriophyllum, but not Isoetes) .

Total invertebrate densities on R. fluitans are not shown in Fig. 3.2 because only ten quantitative samples were collected during the pilot survey and main quantitative sampling program and the lengths of stems were too difficult to measure accurately. The limited information available suggests that invertebrate densities on Ranunculus were less than on the other plants both in terms on numbers/m2 of lake bottom 4 2 (3.2 x 10 ) and numbers/g (dry wt) of macrophyte (5.3 x 10 ). 42

The most valid density units to use depends upon the aims of any investigation (together with practical considerations). For example, if the aim was to determine the total popUlation of a taxon in a particular habitat it would be best to use units such as ~umbers/m2 of lake bottom. This would enable total numbers to be estimated by mUltiplying the density, expressed as above, by the area of habitat. However, for comparison of densities of invertebrates between different species of macrophytes it is better to use units that are based on a property of the particular plant species such as numbers/unit weight of macrophyte, numbers/unit length of macrophyte stem or numbers/surface area of macrophyte. As can be seen in Fig. 3.2 the comparison of densities of invertebrates on the different macrophytes is extremely dependent upon the units used. Numbers/g (dry wt) of macrophyte (e.g., Fig. 3.9) are preferred for comparison of invertebrate communities on different macrophytes since differences in invertebrate abundances between plants are more easily interpreted in these units.

The percentage taxonomic composition of invertebrate communities present on any species of macrophyte is, of course, independent of the units of density when invertebrate densities are numerically based (cf. biomass of invertebrates basis). Mollusca were the most numerous major group on all macrophytes examined (Fig. 3.2), especially on R. fluitans*, but none of the invertebrate communities contained the other major taxa in the same order of rank. Coelenterata formed a higher proportion of the invertebrate communities on E. canadensis (15.4%) (especially at sites in the western sampling area) than on other plants. Crustacea were the only other group to contribute more than 10% to the invertebrate communities on any of the macrophytes (Elodea and Isoetes Fig. 3.2). Insecta comprised a markedly larger percentage of the fauna on native compared with adventive macrophytes (Fig. 3.2 *). The densities of insects/m2 of lake bottom on Isoetes and Myriophyllum were 2.4 times greater than on Elodea, and on Ranunculus only 0.7 times that on Elodea. Densities in terms of numbers of invertebrates/g (dry wt) ofmacrophyte were 6.6 times (Isoetes), 4.0 times (Myriophyllum) and 0.9 times (Ranunculus) the density on Elodea; and in terms of insects/m of plant stem 2.0 times (Isoetes) and 4.4 times (Myriophyllum). The native macrophytes therefore were strongly colonised by insects in Lake Grasmere despite their limited distributions.

* Ranunculus fluitans; Coelenterata 0.7%, Annelida 0.5%, Crustacea 4.1%, Insecta 2.9%, Acarina 5.1%, Mollusca 86.6%. 43

Trichoptera (especially Hydroptilidae) and Chironomidae together comprised 92.0% (Elodea), 97.6% (Isoetes), 94.7% (Myriophyllum) and 94.4% (Ranunculus) of the insects on each of the four species of macrophyte (Appendices 1 and 2.8). Acarina were better represented on macrophytes other than Elodea, but the apparent occurrences of Annelida and Crustacea on plants were probably enhanced by inclusion, in samples, of benthic and planktonic animals respectively

(2) Invertebrate communities at different depths

Quantitative samples of invertebrates on submerged aquatic macrophytes were collected from lake water depths ranging from 1 - 4 m. The numbers of taxa present at each of the four depths are depicted in Fig. 3.3.

(a) Chlorohydra viridissima Cura pinguis Plumatella repens OLIGOCHAETA Chaetogaster SPa CLADOCERA OSTRACODA Eucyclops serrulatus Xanthocnemis zealandica Diaprepocoris zealandiae Paroxyethira hendersoni Paroxyethira tillyardi Triplectides cephalotes Oxyethira albiceps Nymphula nitens CHIRONOMIDAE piona uncata exigua Gyraulus corinna Hydrozetes lemnae potamopyrgus antipodarum Trimalaconothrus novus Sphaerium novaezelandiae (b) NEMATODA Hudsonema amabilis Antiporus strigosulus Oecetis unicolor pycnocentrodes aureola Arrenurus sp. Triplectides obsoleta Physastra variabilis (c) Deleatidium sp. Oecetis iti zelandobius furcillatus (d) Glossiphonia mUltistriata Sigara arguta

(e) Placobdella maorica procordulia gra

Fig. 3.3 Venn diagram of the distribution of invertebrate taxa collected in quantitative samples of submerged aquatic macrophytes from four depths in Lake Grasmere (April 1976 - October 1977). Numbers of samples collected 1 m 94, 2 m 101, 3 m 9 and 4 m 9. 44

Of 37 taxa, 22 were recorded from all depths sampled, but only 17 of these were found on all macrophytes examined (Fig. 3.1). These comprised the 13 species that occurred at every site sampled plus P. repens, P. tillyardi, S. novaezelandiae and O. albiceps. There was a general decline in numbers of taxa with depth. At 1 m, 35 taxa were recorded, including the characteristic of stream-like environments (e.g., Deleatidium sp. and Z. furcillatus) or sandy substrates (oecetis iti) which were unique to this depth. Eight taxa were shared by 1 m and 2 m and two by 1 m, 2 m and 3 m. Only two taxa were not collected in samples from 1 m (p. maorica and P. grayi) and these, relatively rare, species were found only in Elodea samples from 2 m depth. Their apparent absence from other depths may be a consequence of their rarity combined with decreased sampling effort at depths other than 2 m. Nearly 50% of all quantitative samples were collected from this depth (Table 3.6). Despite low numbers of quantitative samples collected from 3 m and 4 m (all in the Elodea zone) I the lower numbers of taxa (22 and 24, Fig. 3.3) collected reflect the true situation as indicated by extensive non-quantitative sampling using a 200 ~m mesh hand-net in these habitats.

Considerable changes in invertebrate density and community composition occurred with depth in the Elodea zone in the western sampling area (Fig. 3.4). Invertebrate communities were dominated by molluscs (1 - 2 m) and Chlorohydra (3 - 4 m) with Crustacea the next most common group. Other salient features were the relatively minor contributions of Insecta and Acarina at all depths (1. 2 - 5.2%) and the relatively similar percentage representation of Crustacea (16.8 - 18.5%) over the depth range 1 - 3 m. C. viridissima densities were highest where Elodea was growing in water 3 m deep (although the Chlorohydra on plant stems could be as close to the surface as 1 m) . The seasonal peak in abundance of Chlorohydra (8 April 1977) was correlated with the maximum Secchi Disc reading (5.0 m) obtained during the study. Increased water transparency and consequent increased light penetration at this time may enhance the production of the symbiotic zoochlorellae (which give Chlorohydra its green colour) and hence boost Chlorohydra abundance. Maximum values of water transparency (Secchi Disc reading ca. 8.2 m) have been recorded in previous years in April and May (stout 1972). Low light penetration (i.e., low Secchi Disc readings) in other months probably limits Chlorohydra abundance. Densities of Annelida, Insecta and Mollusca showed a general decrease with depth as fewer animals with bottom-dwelling habits were included in samples and periphyton levels on 45 plant surfaces diminished. Qualitative differences in periphytic coatings were easily assessed visually or by the slimy nature of the plant surfaces (see also Chapter IV, p. 87).

>­ I- if) x ~ !I ~~ n:: 3 en w x l- if) n:: W (j) 5 w Ol 0'0 ZJa >Z "a c: W . I- a Z"c:

o

j I I I ,. t i • • f 1234 1234 1234 LAKE DEPTH (m)

Fig. 3.4 Total invertebrate density, percentage composition of community in terms of major taxa, and changes in abundance of major taxa between 1 m and 4 m (lake depth) in the Elodea canadensis zone in the western sampling area of Lake Grasmere (April 1976 October 1977). (Numbers of samples: 1 m 9, 2 m = 34, 3 m = 9, 4 m = 9.)

(a) Introduction. Thirty-seven invertebrate taxa were recognised in quantitative samples from 13 sites in the macrophyte zone of Lake Grasmere (Table 3.9). In order to simplify the analysis of the factors affecting associations of invertebrates with macrophyte species and with other habitat characteristics (e.g., substrate, depth and wave exposure) it was necessary to reduce the number of possible comparisons. I decided to use a clustering method of data reduction aimed at aggregating sites with similar invertebrate communities into groups (site groups) and creating species groups consisting of taxa that showed similar site preferences. This treatment decreases the number of comparisons required for determination of factors affecting community 46 composition and (since data from individual sites are combined) provides a more substantial data base for comparative analyses.

Cluster analysis is a numerical method of classification in which entities (such as sampling sites) are grouped according to the similarity of the attributes (for example, species) associated with each entity. In this type of classification, the groups of similar entities are related to one another by a two-dimensional hierarchical tree structure called a dendrogram (e.g., Figs. 3.5 and 3.6). A dendrogram is defined as a diagrammatic illustration of relationship based upon degree of similarity or dissimilarity where the level of branching indicates relative similarity or dissimilarity. In addition to depicting relationship between entities, the order of the entities should correlate with a gradient 0'£ some kind. In a dendrogram such as Fig. 3.5 the order of sites may be correlated with, for example, a gradient of substrate type, wave exposure, or macrophyte growth form.

Many different similarity/dissimilarity measures and/or clustering strategies are available and dendrograms resulting from their use may differ markedly according to the procedures adopted (see Clifford & Stephenson 1975).

The choice of methods is often related to the nature of the data. Ecological data with many zero entries and a few outstandingly high entries suggest transformation (for example, logIO (n + 1», use of the Bray-Curtis dissimilarity measure and a flexible or group average sorting strategy. The final choice of methods depends upon how the data can be interpreted usefully by the biological user and thus to a large extent the 'best' methods to use are those that result in 'sensible' classifications. The effectiveness of a classification may be considered in two ways: the first is whether or not the clusters comprise like grouped with like on the basis of intrinsic data (i.e., data used in dendrogram formation, such as species' abundances and distributions) and the second depends upon the extent to which groupings obtained by intrinsic data reflect extrinsic attributes (e.g., environmental features).

(b) Data processing. For the purposes of computer analyses, five species (Plumatella repens, Placobdella maorica, Procordulia grayi, Triplectides obsoleta and Oecetis iti) were not considered since they were either not quantified (P. repens) or were represented in quantitative samples by only one or two specimens (Appendix 2). Five further taxa (Oligochaeta, Nematoda, Cladocera, ostracoda and 47

Chironomidae) were not identified to species due to taxonomic short­ comings at the time of the analyses.

A species X sites matrix or two-way table' (similar to Table 3.9; containing species densities as numbers of invertebrates per sample summed over all sampling dates) was constructed using information collected during the quantitative sampling program (September 1976 - October 1977) (Appendix 2). To facilitate cluster analyses, because attribute values (i.e., individual invertebrate species' abundances) covered a wide range (between 0 and about 400 individuals per sample, Table 3.9) this matrix was transformed using a 10g10 (n + 1) transforma- tion. Dendrograms depicting the degree of association of site groups (Fig. 3.5) and species groups (Fig. 3.6) were derived with the aid of Professor W. Stephenson (University of Queensland, Australia) using the Bray-Curtis dissimilarity measure and a group fusion sorting strategy.

30 - I

I 1 - IA B C 1D I

I

,--'--

o E@1. E@2. EQ) E@ E@. N@2. W@2 W@3. W@4, S@. so), W@1. WCD· SITE

Fig. 3.5 site group dendrogram. (All quantitative samples combined, September 1976 - October 1977).

The Bray-Curtis measure, which is considered desirable where differences in species dominance and zeros (i.e., species absence) are involved, defines the dissimilarity of two sites (or individuals) as

n E x Ij - X lj

n ) E Xl + XZj 1 j 48

where n is the number of attributes (for example, species) x and 1j X j are the values (for example, sites) (Clifford & Stephenson 1975). 2 Various attempts have been made to replace, with tests of significance, personal judgments of the results of classification. These tests, which are of four main types, are open to criticism on theoretical grounds, but in practice tend to confirm common sense judgments (see Clifford & stephenson 1975, for more complete discussion). The 'pseudo F test' (Prof. W. stephenson, pers. comm.) was used to find out which species 'conformed noticeably' (see Stephenson, Williams & Cook 1974) to their site groups by applying the test to the rows of a species X sites two-way table. species that 'conform noticeably' to their site groupings usually are characterised by marked differences in numerical dominance and/or constancy of occurrence in different site groups.

(c) Site groups. The associations of sites, in terms of their invertebrate communities are depicted in the site group dendrogram (Fig. 3.5). The aim of this cluster analysis was to aggregate sites with similar invertebrate communities into groups in order to simplify the determination of factors affecting community composition. A compromise is necessary between excessive clumping (leading to over­ simplification and loss of information) and no aggregation at all. Arguments can be presented favouring various numbers of site groups. For example, six site groups could be chosen [viz., (E Eland E E 2), (E I), (E M and E R), (N E 2, WE 2, WE 3 and W E 4), (S M, S I and WEI) and (W I)] since E I and W I are the two single sites whose invertebrate communities differ most from those at any other sites. In this respect they may deserve 'group' status. However, environmental conditions (for example, wave action and sediment type) at these two sites (Table 3.5) are not sufficiently different from those at the other sites in groups A and D respectively (Fig. 3.5) to warrant separation. Consequently, four site groups (A, B, C, D) were chosen using knowledge of both the environmental features (Table 3.9) and invertebrate community composition (Table 3.8) at the sites to determine the component sites of each group.

Physical features of the site groups. The physical features of the study area have already been considered (Chapter II) (Table 3.9) and the following discussion emphasises the main points in the light of the composition of the site groups. 49

Group fusion dissimilarity (Bray -Curtis) o 50,. I I I I I I I

Sigara. arguta , Hudsonema amabilis I r 1 I Physastra variabilis

Glossiphonia multistriata I I 2 Diaprepocoris zealandiae

Pycnocentrodes aureola

Nymphula nitens J-l

Triplectides cepha.lotes 4 Paroxyethira tillyardi 5

Trimalaconothrus novus

Chlorohydra viridissima 6 OLlGOCHAETA

Paroxyethira hendersoni

Chaetogaster sp.

CHIRONOMIDAE

CLADOCERA ;J-~ f- Gyraulus corirna r-- Cura pinguis

Eucyclops serrula.tus

piona uncata exigua ~ r- Hydrozetes lemnae

Potamopyrgus antipodarum 0

NEMATODA OSTRACODA ~7 Sphaerium novaezelandiae

Xanthocnemis zealandica 8 Oxyethira albiceps I 9 Oecetis unicolor

Arrenurus sp~

Antiporus strigosulus 1 10 Deleatidium sp. I I Zelandobius furcillatus I

Fig. 3.6 Species group dendrogram. (All quantitative samples combined, September 1976 - October 1977). 50

Site A

The three sites (E E 1, E E 2 and E I) comprising this group are on the eastern side of Lake Grasmere (Fig. 2.1) where the substrate ranges from sand and stones to large rocks and boulders. Prominent features are the inputs of branches and leaves from overhanging mountain beech trees (Nothofagus solandrii var. cliffortiodes) and the shading effect of the adjacent hillside and trees, especially when the sun is low in winter. Wave action is weak to mode.rate.

Site group B

This group comprises sites E M and E R and shares the physical features of site group A. The important feature here is the very restricted distribution of the macrophytes Myriophyllum propinquum and Ranunculuq fluitans. In this area of the lake they occured only in a very narrow band (less than 1 m wide) between the Isoetes and Elodea zones in water about 1 m deep.

Site C

This relatively homogeneous group comprised four sites (N E 2, W E 2, W E 3 and W E 4) within an extensive monoculture of Elodea canadensis and represented the dominant macrophyte habitat in the lake (about 90% of the macrophyte-covered area) . Wave action is only moderate to weak since the water is deeper (2 - 4 m) and the substrates are highly organic, comprising silt and fine particles resulting from the decomposition of the macrophytes. sites within this group have more similar invertebrate communities than those comprising any other site group. N E 2 and W E 2 were the two least dissimilar sites sampled (Fig. 3.5), which is not surprising since they are at the same depth (2 m) in the same extensive Elodea monoculture (Fig. 2.1).

Site group D

This group is the most heterogeneous of the four site groups chosen. This heterogeneity is reflected in both the environmental features of the sites (Table 3.5) and in the low degrees of dissimilarity between sites in the site group dendrogram (Fig, 3.5). The sites are all shallow, being in water less than 1 m deep, with moderate to strong wave action especially during prevailing north- westerly winds. The two southern sites (S I and S M) are influenced also by various inlet streams arising from springs in swampy ground 51 near the lake edge (especially Inlet A, Stout 1972 Fig. 1) (Fig. 2.1). The substrates range from mud (western sites) to stones (southern sites) with large accumulations of organic material, mostly derived from macrophyte decomposition. At the southern end of the lake in the zone subjected to strong wave action (i.e., in water less than 1 m deep) macrophyte growth is relatively sparse and found in clumps act as sediment and organic matter traps. Between these clumps, finer sediments may be eroded away leaving a more stony bottom. The lake bottom in this region of the littoral zone is relatively uneven comprising small humps of macrophytes with hollows between. The conditions in these zones are more reminiscent of lotic than most lentic situations.

Discussion. The site group dendrogram . 3.5) highlights the marked difference between invertebrate,-macrophyte communities of the eastern sampling zone (site groups A and B) and communities in other areas of the lake groups C and D) • Site group B comprises communities whose occurrence was restricted to a very narrow band along the shore, although it is otherwise surprising that these sites do not show greater similarity to site EEl from site group A since this site is in very close proximity to the sites of group B. Site group C, on the other hand, represents about 90% of the macrophyte-covered area of the lake and is the most homogeneous group, being virtually a monoculture of Elodea canadensis. Site group D, the most heterogeneous assemblage, is characterised by highly organic substrates and marked development of periphytic coatings on plant surfaces.

An environmental gradient which may be correlated with the order of site aggregation (i.e., EEl, E E 2, •••.. W E 1, W I as on Fig. 3.5) is the degree of organic matter accumulation in the substrate and on plant surfaces. The eastern sites (site groups A and B) are characterised by relatively 'clean' macrophyte surfaces with only moderate periphytic development (due perhaps to a shading effect of the hillside) and substrates that are sandy/stony with very coarse particulate organic matter (beech leaves, twigs, etc.). Site group C has relatively silty substrates with high organic content (decomposing macrophyte material) whereas sites in group D have a variety of substrate types with macrophytes notable for their relatively stunted growth (due perhaps to wave action and shallowness of the water) and very thick coatings of periphyton and organic debris. The above gradient to some extent is due to the effect of wave action. Site groups A and B probably represent 'erosional'sites where fine organic material is removed from 52 the substrate by the moderate wave action, whereas sites in group D, which are subjected to the strongest wave action, are net 'depositional' sites due to the effects of the prevailing northwesterly winds. Sites in group C on the other hand are not subjected to strong wave action as they are in deeper water and consequently there is not likely to be any removal or addition of organic material to or from other areas. The substrates here are primarily the result of in situ decomposition of Elodea canadensis and organic matter processing by invertebrates.

(d) Species groups.

and their occurrence at the site

Introduction. The purpose of this section is to outline the taxonomic composition of the species groups and to account for this composition using knowledge of the habitat requirements of the invertebrate species (see p. 34) and distributional information from sampling invertebrate-macrophyte communities in Lake Grasmere (Table 3.9, Appendix 2).

Because clustering strategies are 'group-size dependent' (Stephenson, Williams & Lance 1970), groups that differ greatly in size should not be selected by defining a fixed level of dissimilarity. Instead, groups are best selected by following down the dendrogram until the species remaining in a group subjectively have similar distributions among the site groups. In other words, species groups should be selected so that they make ecological sense.

In this manner, ten species groups (numbered 1- 10, Fig. 3.6) were selected from the species group dendrogram. Knowledge of each species' distribution, from data obtained during the quantitative sampling program (Table 3.9 and Appendix 2), was used to delimit species groups.

The representation of each species group at each site group (Tables 3.10 and 3.11) was determined by calculating the mean number of individuals of the species group per sample at the site group and then expressing this as a percentage of the sum of values for all site groups. The percentage representation was adjusted to give equal weight to each site group, since different numbers of samples were taken from each (Table 3.17). Thus, the site group that has the highest mean numbers of a species group per sample has the highest percentage 53

occurrence of that species group. If the percentage occurrence had been calculated using total numbers of individuals collected from each site group, the percentage representation of a species group would be biased in proportion to sampling effort. This would tend to obscure any preferences that certain species groups may have for different macrophytes, substrates, sampling areas of other environmental features and also make interpretation of the dendrograms, which were derived on a mean numbers per sample basis, more difficult.

Table 3.10 Percentage representation of species groups (by mean numbers of individuals per sample) at site groups, and the percentage of total invertebrate numbers per sample occurring at each site group (September 1976 - October 1977).

- = not present.

Species groups Site groups

A B c D

I 39.0 51. 8 2.3 6.9 2 49.3 29.4 21. 3 3 18.1 45.0 36.9 4 8.4 12.1 79.4 5 38.6 52.3 2.8 6.3 6 26.8 12.9 32.1 28.2 7 21.6 2.3 3.6 72.5 8 48.6 51.4 9 3.0 1.5 95.5 10 53.6 46.4

% Total numbers 26.8 13.1 31.4 28.7

Species group I

This species group comprised the hemipteran Sigara arguta, the trichopteran Hudsonema amabilis, and the gastropod mollusc Physastra variabilis. These species seem to have little in cornmon. Although they are all distributed widely (although not necessarily abundantly) throughout New Zealand this is a feature shared by many species in other groups and does not explain their association here. H. amabilis and 54

P. variabilis 'conformed noticeably' to site groups A and B (P < 0.01% pseudo F test) although the species group was otherwise relatively weakly clustered (Fig. 3.6) and most common at site group B (Table 3.10). The presence of sandy/stony/rocky substrates (rather than very fine organic sediments), the steeply shelving lake shore, and the relatively weak wave action may be key factors in their occurrence. These species were most common in shallow water: S. arguta in sheltered areas; H. amabilis (on Myriophyllum and Ranunculus and especially mineral substrates) and P. variabilis both in the eastern sampling area. H. amabilis was the third most abundant caddisfly in the macrophyte zones of the lake (after the two paroxyethira species) but P. variabilis was outnumbered by potamopyrgus antipodarum by at least 1,200 to one (Table 3.9, Appendix 2).

Species group 2

This species group comprised two weakly clustered species; the leech Glossiphonia multistriata and the hemipteran Diaprepocoris zealandiae, neither of which 'conformed noticeably' to any site groups. G. multistriata has been recorded previously on stones and macrophytes in Lake Grasmere (Stout 1977) and D. zealandiae was the more common of the two corixids in the lake (Table 3.9). A common factor linking these two species in Lake Grasmere seems to be an association with Elodea canadensis. Ninety-two percent of G. multistriata and 75% of D. zealandiae collected during the quantitative sampling program were associated with this macrophyte from site groups A and C (Table 3.9). This species group was most common at site group A and present also at site groups C and D (Table 3.10).

3

pycnocentrodes aureola (Trichoptera) and Nymphula nitens (Lepidoptera) were associated at a high level of dissimilarity (Fig. 3.6) suggesting that their association is relatively weak. Neither 'conformed noticeably' to any site groups. P. aureola was not common on macrophytes in Lake Grasmere (Table 3.9) but was very common in shallow water on mineral substrates. Conversely, N. nitens, during 1976 and 1977, was found most often associated with shoot tips of Myriophyllum (Table 3.9). In 1978, however, when densities of N. nitens increased dramatically (see Chapter V) many more larvae were found associated with other macrophytes such as E. canadensis. This species group was most common in site groups B, D, and A (Table 3.10) with preference for si tes where macrophyte stem lengths were short (0. I - 0.3 m) (Table 3.9). 55

4

Triplectides cephalotes (Trichoptera) was the sole member of this species group and 'conformed noticeably' to site group D (P < 0.01, pseudo F test). It was present also in site groups C and B (Table 3.10) . These records earlier observations on preference for shallow water sites. Cowley (1978) recorded larvae living amongst stones on a rocky beach in Lake Grasmere, although my observations suggest that a proportion of the population in the lake is associated with aquatic macrophytes.

5

The sole occupant of this species group, the hydroptilid caddisfly paroxyethira tillyardi, ,'conformed noticeably I to eastern sites (site groups A and B) (P < 0.01, pseudo F Nearly 90% (in terms of mean numbers per sample) were collected from these areas (Table 3.10) and those collected from other areas preferred shallow sites (Table 3.9). A distinct preference for sandy/stony substrates rather than silty/muddy substrates was evident.

Species group 6

The group comprised thirteen taxa (Table 3.11) whose common feature was their occurrence at all sites sampled. Six taxa 'conformed noticeably' to even representation at all site groups

(Oligochaeta (P < 0.05), Chaetogaster sp. (P < 0.01), Cladocera (P < 0.001), Eucyclops serrulatus (P < 0.05), Chironomidae (P < 0.001), and Gyraulus corinna (P < 0.001».

Species group 6 contains the eleven most common taxa (by total numbers collected during the quantitative sampling program) and the 13th and 16th ranked species and comprises species that are characteristic of macrophyte zones of lakes (see Winterbourn & Lewis 1975). This group makes up over 95% (by numbers) of the fauna at each site group (see Tables 3.17 and 3.18 and associated discussion).

P. antipodarum was the most common macro invertebrate at almost every site (Table 3.9) averaging over 50% by numbers in all quantitative samples combined. C. viridissima, the second most common species, was especially common on E. canadensis in site group C (Tables 3.9 and 3.11). There was a marked seasonality in its occurrence (see Figs. 3.9a, 3.10- 3.13) and marked change in abundance with depth (Fig. 3.4). The third most abundant species, the copepod E. serrulatus, was most common also at site group C (Table 3.11) where reduced wave action probably enhanced 56 its abundance. G. corinna, the fourth most common macroinvertebrate on macrophytes in Lake Grasmere, was found also on rocky and sandy substrates. On plants it was present mostly at site groups C and D (Table 3.11).

Table 3.11 Percentage occurrence of the taxa of species group 6 among the four site groups (on a mean numbers per sample basis).

Site group Taxon A B C D

Trimalaconothrus novus 18.4 6.1 17.7 57.8 Chlorohydra viridissima 20.5 8.1 67.8 3.6 OLIGOCHAETA (except Chaetogaster) 13.9 0.7 44.1 41. 3 paroxyethira hendersoni 8.6 17.8 13.7 59.9 Chaetogaster sp. 4.0 8.9 38.5 48.6 CHI RONOMI DAE 11.1 9.4 17.8 61. 7 CLADOCERA 7.8 5.7 48.7 37.8 Gyraulus corinna 8.7 2.2 48.2 40.9 Cura pinguis 30.1 12.4 14.1 43.4 Eucyclops serrulatus 25.0 7.4 43.5 24.2 Piona uncata exigua 34.6 19.8 34.1 11.5 Hydrozetes lemnae 23.1 29.9 14.6 32.4 Potamopyrgus antipodarum 34.5 16.6 20.7 28.2

Cladocera from quantitative samples were not identified to the specific level prior to April 1977. However, seven species were identified from samples collected between April and October 1977 inclusive. During this period cladoceran species composition (on a mean numbers per sample basis; all sites combined) was as follows: Bosmina meridionalis 40.1%, Graptoleberis testudinaria 36.9%, Alona guttata 17.1%, Chydorus sphaericus 6.3%, Ceriodaphnia dubia 0.6%, Ilyocryptus sordidus 0.2%, and Neothrix armata 0.1%. Leydigia ?australis and Simocephalus vetulus occurred also but were not collected in quantitative samples.

Distinct differences in species composition of Cladocera occurred in each site group (Table 3.12). G. testudinaria was most abundant at site group C. These two species constituted nearly 85% of cladocerans collected and A. guttata and C. sphaericus were the 57 only other species to contribute more than 10% by numbers at any site groups between April and October 1977.

Table 3.12 Species composition (%) of Cladocera in quantitative samples from each site group and total numbers of each collected (April - October 1977).

Site group Species Total numbers collected A B C D

Bosmina meridionalis 1168 20.5 67.3 9.0 Graptoleberis testudinaria 763 53.9 75.0 29.2 41. 3 Alona guttata 240 20.5 1.0 38.2 Chydorus sphaericus 99 5.2 25.0 2.2 10.1 Ceriodaphnia dubia 12 0.4 0.9 Ilyocryptus sordidus 2 0.4 Neothr ix armata 1 0.2

Cladoceran densities (Table 3.13) were much lower in the eastern sampling area (site groups A and B) than elsewhere. There was clearly a greater planktonic influence on the phytomacrofaunal communities at site group C. Indicative of this influence was the prevalence of the planktonic B. meridionalis at site group C and the presence of C. dubia exclusively at site groups C and D (although only a few individuals were collected, Table 3.12) (Table 3.13).

Table 3.13 Percentage representation of each cladoceran species (by mean numbers per sample) between the four site groups, and overall mean numbers per sample (all species combined) at each site group (April - October 1977).

Species Site group A B C D

Bosmina meridionalis 2.4 88.5 9.1 Graptoleberis testudinaria 6.8 5.9 41.6 45.7 Alona guttata 5.6 3.0 91.4 Chydorus sphaericus 3.8 11.5 18.6 66.1 Ceriodaphnia dubia 36.7 63.3 Ilyocryptus sordidus 100.0 Neothrix armata 100.0 mean numbers/sample 3.9 2.4 44.0 34.1 58

G. testudinaria, A. guttata, and C. sphaericus are all weed dwelling species with distributions biased towards site groups 0, or C (Table 3.13) • These areas (cf. site groups A and B) had more available macrophyte substrates and/or greater quantities of organic material and periphyton amongst or adhering to the stems of macrophytes. It is likely, therefore, that the distributions of the weed dwelling cladocerans were in response to available food or attachment sites.

Besides Chaetogaster sp., four oligochaetes were recorded from macrophyte zones of Lake Grasmere (Table 3.8), although they were more common in mud than on macrophyte surfaces. Although oligochaetes from quantitative samples were not identified to species on a regular basis, Lumbriculus variegatus and Chaetogaster sp., and to a lesser extent, Eiseniella tetraedra were the dominant species on macrophytes. The Oligochaetes (including Chaetogaster sp.) were most common in site groups C and 0 (Table 3.11) where the presence of large quantities of organic detritus amongst the macrophytes was likely to be the main influence upon their distribution.

Chironomidae were not identified specifically during original processing of quantitative samples due to taxonomic difficulties and, therefore, were lumped in the dendrogram analyses. Subsequently, further identifications have been possible (Stark in press) and ten taxa have been recognised (see also Chapter VI, Appendix 3 and Table 3.14) .

Table 3.14 Species composition (%) of Chironomidae in quantitative samples from each site group and all sites combined (September 1976 - October 1977).

Total Site group Taxa numbers collected A B C 0 Overall

Cricotopus spp. 761 43.7 91.9 87.9 64.5 70.4 Macropelopia/Gressittius 147 45.2 10.3 9.5 13.5 Tanytarsus vespertinus 98 16.9 9.1 Orthocladiinae A 27 4.7 2.5 Ablabesmyia mala 25 8.7 5.4 0.3 1.9 2.3 Orthocladiinae B 10 2.7 1.6 0.9 Chironomus zealandicus 6 0.8 1.2 0.2 0.6 Orthocladiinae C 4 0.7 0.4 Pentaneura sp. 2 1.6 0.2 Syncricotopus pluriserialis 1 0.3 0.1 59

Distinct differences in chironomid species composition occurred

in each site group (Table 3.14). Cricotopus (perhaps two species) was the dominant chironomid at three site groups (B, C and D) and in

terms of overall abundance. At the remaining site group (A), the abundance of Cricotopus spp. was similar to that of Macropelopia/ Gressittius (two? species). All these taxa comprised nearly 85% of chironomid larvae collected during the quantitative sampling program. T. vespertinus was the only other species representing more than 10% of chironomid numbers at any site group and was found exclusively at site group D.

Table 3.15 Percentage representation of each chironomid species (by mean numbers per sample) in the four site groups, and overall mean numbers per sample (all species combined) at each site group (September 1976 - October 1977).

site group Taxa A B C D

Cricotopus spp. 7.0 12.7 23.0 57.3 Macropelopia/Gressittius 39.1 14.7 46.2 Tanytarsus vespertinus 100.0 Orthocladiinae A 100.0 Ablabesmyia mala 35.8 19.1 2.0 43.1 Orthocladiinae B 20.0 80.0 Chironomus zealandicus 21.7 52.2 26.1 Orthocladiinae C 100.0

Pentaneura Spa 100.0 Syncricotopus pluriserialis 100.0 mean numbers/sample 2.7 2.3 4.4 14.8

Chironomid densities were much higher site group D than elsewhere (Table 3.15) due to greater species richness resulting from increased habitat heterogeneity and the inclusion of bottom-dwelling animals in samples. All but one (C. zealandicus) of the species occurring at site group D was most common at this most heterogeneous

group. Cricotopus spp. favoured site groups C and Df whereas the predatory Tanypodinae (especially Macropelopia/Gressittius and A. mala) were most abundant at shallow sites (site groups A and D) (Table 3.15). C. zealandicus, a species common in the benthos below the macrophyte 60 beds of Lake Grasmere (Timms in prep.), was infrequently collected from macrophytes but was most represented at site group C (the E. canadensis 'monoculture').

Three species of Acarina were included in species group 6. Hydrozetes lemnae was well represented at all site groups (Table 3.11) but had highest densities at shallow sites in site groups A and D (Table 3.9), Piona uncata exigua was most common at site groups A and C (Table 3.11), especially at sites characterised by E. canadensis (Table 3.9), Lowest densities were recorded from shallow sites with moderate to strong wave action (mainly site group D) • This distribution is consistent with the known planktonic/littoral habits of Piona (Stout 1969b,1976, Winterbourn & Lewis 1975). The third species, Trimalaconothrus novus, was very common in clumps of I. alpinus, especially near the leaf bases. In July 1976 white nymphs were found inside a decomposing Isoetes stem indicating that such sites may be used for egg-laying. T. novus was most common at site group D (Table 3.11).

Paroxyethira hendersoni was the most common trichopteran living in Lake Grasmere (Table 3.9) and the most frequently taken as adults in hand net (Appendix 5.3) and light trap (Appendix 5.4) collections from the lake shore. It was best represented at site group D (Table 3.11) I especially at sites W I and S M (Table 3.9). There appeared to be a distinct preference for zones of I. alpinus and M. propinquum that occurred in shallow water.

The remaining member of this species group is the flatworm Cura pinguis. This species was distributed widely in Lake Grasmere but was most common at site groups D and A (Table 3.11) and especially at sites W If EEl and E E 2 (Table 3.9).

Species group 7

This species group comprised three taxa characteristic of benthic habitats. All 'conformed noticeably' to site groups A and D (Nematoda, pseudo F test, P < O.Oli Ostracoda, P < 0.01; Sphaerium novaezelandiae, P < 0.05).

Nematoda and Ostracoda (except for the very distinctive Gomphocythere duffi) were not identified to species during the quantitative sampling program due to taxonomic shortcomings. However, six species of Ostracoda (Table 3.8) have been identified subsequently (Dr M.A. Chapman, pers. comm.). All are characteristic of benthic (Scottia insularis, Herpetocypris pascheri, Prionocypris marplesi, 61

G. duffi and .Darwinula repoa) and macrophytic (Cypridopsis vidua and D. repoa) habitats (Chapman 1976). By far the most common species in samples was D. repoa. s. insularis (a small dark green that occurred in large numbers in highly organic substrates near the swampy inlet stream, but non in quantitative samples) and P. marplesi have hardly ever been recorded since they were described by Chapman (1963) (Dr M.A. Chapman, pers. comm.).

S. novaezelandiae was present in all sandy to muddy habitats in

I the macrophyte zones of Lake Grasmere and it occurs in the benthos as well (Jamin 1976, Timms in prep.).

Species group 7 was most common at site groups 0 and A (Table 3.10). There was a trend to decreasing abundance of these. taxa with depth (or increasing height of macrophyte) as the bottom faunal influence on samples decreased. This difference was marked especially in the western sampling zone: W I was the most preferred site (with the shortest macrophyte), followed by WEI, with a marked reduction in representation where macrophyte stems were longer, in deeper E. canadensis zones (W E 2 - W E 4) (Table 3.9) of site group C.

Species group 8

The larva of the common red damselfly, Xanthocnemis zealandica, was the sole member of this species group and did not 'conform noticeably' to any site groups. X. zealandica was not common in Lake Grasmere during most of 1977 and was collected rarely in quantitative samples An indication of the difference in its abundance between 1976 and 1977 is· given by Fig. 5.13 (p. 134) where similar effort with a hand-net yielded approximately seven times as many larvae in the last four months of 1976 as in the same period the following year. The extensive northern and western zones of E. canadensis group C) were the favoured habitat in the lake (Table 3.9) and it was from these areas that animals were collected most readily for life history information and faecal analyses (Chapters IV and V) . In terms of site groups ,. X. zealandica was represented almost equally in groups C and 0, the latter due entirely to site WEI (Tables 3.9 and 3.10).

Species group 9

Oxyethira albiceps (Trichoptera: Hydroptilidae), Oecetis unicolor (Trichoptera: Leptoceridae) and Arrenurus Spa (Acari: Prostigmata: Arrenuridae) comprised this species group but did not 'conform noticeably' to any site groups. 62

o. albiceps was collected in quantitative samples only from western and southern areas (site group 0) although it was taken occasionally in hand-net sweeps from deeper E. canadensis zones in the western sampling area (site group e), especially when the biomass of green filamentous algae was high (e.g., October 1977). Highest densities, however, were found on algal covered stones at the southern end of the lake.

O. unicolor was collected in quantitative samples from only two sites (W Eland W I, Table 3.9) in site group O. In addition, it was present on sandy substrates in the southern zone, and in greatest densities near the outlet stream at the northern end of the lake where substrates were sandy and the water less than 1 m deep.

The undescribed species of Arrenurus was best represented at sites WEI and W I (site group 0) but was present also at E E 2 (site group A) and W E 2 (site group e). The known New Zealand species of Arrenurus (A. rotoensis and A. lacus) have been recorded only from

ponds and small lakes (Stout 1953~but little is known of their ecology.

Species group 9 was characteristic of site group 0 although it is likely that macrophyte zones are not the preferred habitats since O. albiceps was most abundant on algal covered stones, O. unicolor on sandy substrates and Arrenurus sp. also probably has benthic habits.

Species group 10

The tenth species group comprised the dytiscid beetle Antiporus strigosulus, the mayfly Deleatidium sp. and the stonefly Zelandobius furcillatus. Only the first named 'conformed noticeably' to any site group (site group A, P < 0.05, pseudo F test).

A. strigosulus was not abundant in Lake Grasmere and was recorded (as one adult and one larva) in quantitative samples at two sites (Table 3.9) in site group A. Non-quantitative hand-net collections suggested that A. strigosulus was present in shallow habitats in the eastern area of the lake, occasionally near the outlet stream in the northern area and in the shallow sites in western and southern areas of the lake. It is an active swimmer and its abundance would almost certainly be underestimated using the quantitative sampler.

Only three specimens of Deleatidiumsp. were collected in

quantitative samples, from two sites in site group 0 (8 I and WEI) and one in site group A (E I) (Table 3.9). Greatest numbers were present along eastern and southern stony shores of Lake Grasmere where 63 the wave action and substrates created conditions that were more typical of streams. Deleatidium could not be regarded as a permanent member of the invertebrate communities on aquatic macrophytes in this lake.

Seven specimens of Z. furcillatus were collected in quantitative samples from two sites in site group D (S I and S M) and one in site group A (E I) (Table 3.9). All three of these shallow water sites had stony substrates and were subjected to wave action.

Species group 10 was represented on at site groups A and D and mostly at sites with stony substrates and moderate to strong wave action.

3.3.5 Seasonal Changes in Invertebrate Communities at the Site Groups

(1) Introduction

The following discussion is concerned with seasonal changes in community structure at each of the four site groups in general terms only. As stated earlier (p. 21) it was anticipated that, since logistic considerations prevented very intensive sampling, the data obtained could not be subjected to detailed month-by-month analysis. Therefore, a variety of indices of species diversity and community similarity are used here in an attempt to examine general seasonal trends in the relative abundance of the taxonomic groupings of Coelenterata, Annelida, Crustacea, Insecta, Acarina and Mollusca and seasonal trends in overall community composition at each site group particularly in relation to any marked changes in between-site comparisons. More detailed of the seasonality of certain species are described in Chapter V.

(2) Procedure

(a) One of the measurable characteristics of any collection of organisms is its 'diversity'. Diversity in this context refers to the degree of uncertainty attached to the identity of any randomly selected individual from a collection of organisms. A collection in which all individuals belong to one species has no diversity whereas one in which every individual belongs to a different species has maximum diversity because it is impossible to the identity of the next species. A variety of indices have been derived to measure diversity (Pielou 1966).

diversity has two components: (i) the number of (i.e., species richness) and (ii) equitability or evenness of the 64

allotment of individuals among the species. Thus, both a greater number of species and a more even or equitable distribution of individuals among the species will contribute to increaseQ diversity.

A commonly used diversity index is that derived independently by Shannon and Wiener (Shannon & Weaver 1949)

s H' - E Pi log i==l

where H' is the population value of the average diversity of a sample from an indefinitely large population; s is the number of species, and Pi is the proportion of the ith species in the population.

Since species diversity has two components it is of interest to know how each contributes to total diversity, To this end we may also calculate, separately, indices of species richness and species evenness.

Species richness (although in its simplest form the number of species) can be calculated from

S d log N where S == the number of species and N = the number of individuals (i.e., a measure of sample size) (Odum, Cantlon & Kornicker 1960, Whittaker 1975). This index Whittakers index of species richness) corrects for the effects of sample size since a larger sample may be expected to contain a greater number of species than a smaller one.

The index of species evenness used here was derived by Pielou (1966) , H' H' J' = H'max. log S

In this case, species evenness is defined as the ratio of observed species diversity (H') to maximum species diversity (H'max. == log S).

(b) indices. Two community similarity indices were used to determine the times of year when most change occurred in the composition of the animal communities at the different site groups. These indices were the coefficient of community (CC) and the percentage similarity of community (PSc) (Jaccard 1932, Whittaker & Fairbanks 1958, Johnson & Brinkhurst 1971, Barton & Hynes 1978). CC measures the percentage of species shared by two samples, community types or habitats among the total number of species represented in both. 65

c Thus, CC in which a is the number of species in the first a+b-c sample, b the number in the second sample and c the number in common.

The second index, one of the most widely used indices of its type, compares animai communities by the numbers of individuals of each species shared by the two samples, communities or habitats.

PSc = 100 - 0.5 E la-bl == E min (a,b) where a and b are, for a given species, the percentage of samples A and B which that species represents.

By calculating CC and PSc values between adjacent sampling dates one can determine when there is the greatest change in the composition

of the animal communities. The lower the values of CC or PSc (i.e' l the lower the degree of similarity) the greater the change between samples. In addition, by applying the indices to replicate samples (i.e., samples that are assumed to be very similar: Cc and PSc should approach 100%) one can not only test sampling replicability, but also set up criteria to assess degrees of affinity between samples. Johnson & Brinkhurst

(1971) considered that high affinity between stations (= samples) was indicated when the values of CC and PSc between stations did not fall below the lowest values of the indices within stations, as assessed from the values obtained from the comparison of replicate samples.

CC measures relative similarity in terms of species composition and may overvalue minor species to the neglect of differences in dominance. This index is ~ery simply applied, however, because it requires only presence/absence data. This simplicity is also its major disadvantage because abundant and rare species are given equal weight. PSc, by contrast, measures relative similarity of numerical composition in terms of species populations and usually leads to grouping of communities by dominants or most abundant species. Its weakness may be in overvaluing the sharing of dominants to the neglect of differences in overall community composition, which is the result also of the presence of rarer species. Used together the two indices are of more value than any single index. comparing their relative values it is possible to determine, for example, whether high affinity of samples is due to the sharing of most species and/or to the occurrence of species in about the same proportions.

(3) Results

(a) Species diversity. Species diversity tended to be lower in eastern sampling zones (site groups A and B) than in other areas (Fig.3.7a). 66

10 ~ >­ b I- (f) (f) (f) w 0::: z W ::r: > u °0.5 0: (f) w (f) ~ u u W W 0... 0... (f) (f)

0

30 1.0 c d :J ". 0. o ~.:...... •...•.. -...... w 0::: (f) lii 10 x' ~

::E 'Jl> • ~ ••••• • ·m u ::::> W z 0... (f)

0+A~~~~~~~~~~~~~~~ O~~~~~~~r.=~~~~~~~~~ A SON 0 J F M A M J J A S 0 ASONDJFMAMJJASO 1976 1977 1976 1977

Fig. 3.7 Seasonal changes in (a) species diversity, (b) species richness, (c) number of taxa, and (d) species evenness for each site group during the time of the quantitative sampling program.

In general, diversity was highest during the summer and late autumn- winter at most site groups. site group A showed the least seasonal change in diversity, It did not exhibit the prominent decrease in April 1977 (Fig. 3.7a) shown by the other site groups, a decrease due to dramatic changes in P. antipodarum (especially at site groups Band D) and C. viridissima (especially at, site group C) (see also Figs. 3.9a and f).

Whittaker's index of species richness suggested there was relatively little change in species richness during the sampling period (Fig. 3. 7b). Site group D, the most heterogeneous group, always had the highest species richness followed, on average, by site groups A, Band C in that order. This order reflected the order of site group heterogeneity (see discussion on physical features of site groups). On a numbers of taxa basis (which is another way of expressing species richness) the series of decreasing species richness at the site groups was D, A, C, B. This series did not reflect the order of site group heterogeneity quite as well as Whittaker's index although seasonal variations were more apparent (Fig. 3.7c). Sampling of the habitats 67 of site group B, which were very restricted in distribution, was difficult and fewer samples were collected than from the other site groups, especially after April 1977 (Table 3.16). The number of taxa

Table 3.16 Numbers of samples collected from the site groups (September 1976 - October 1977).

Site Sampling date group 2/9 2/11 2/12 20/1 2/3 8/4 10/5 20/6 13/7 3/10 Total

A 6 6 6 3 6 5 5 5 0 5 47

B 1 3 4 1 2 2 1 1 0 1 16 C 8 8 8 8 8 8 8 8 8 6 78 D 4 6 4 4 5 4 3 4 5 0 39

collected (Fig. 3.7c) from site group B correlated well with the numbers of samples collected (Table 3.16). This suggests that Whittaker's index of species richness (which accounts for the influence of sample size variations on richness) provided a more realistic assessment of the species richness of this habitat.

The second component of species diversity (i.e., evenness) showed seasonal variation (Fig. 3.7d) that was almost identical in pattern to the changes in total diversity. This indicates that the seasonal pattern of species diversity was due primarily to changes in species dominance rather than in number of species present.

(b) Community similarity. CC and PSc values between adjacent sampling dates were calculated for each of the site groups and plotted midway between sampling dates (Fig. 3.8). For site group A CC and PSc had similar numerical values. There was least change in community composition during autum early winter and most change in November December. The decline in CC and PSc evident between June and October 1977 probably was mostly a function of the length of this time interval (since, even with a constant rate of seasonal change, more change might be expected to occur over a longer period). For most of the year, PSc was higher than CC suggesting that there was more change in species dominance than species presence or absence. Comparison of the community composition on 3 October 1977 with that on 2 September and 2 November 1976 (Fig. 3.8A) suggested there was marked similarity in abundance of 68

100 A 8 ..

% 50

100 o

% 50 ..

0+7T7~~~~~~~~~~~~"-~ A SON 0 J F M A M J J A S 0 A SON 0 J F M A M J J A 5 0 1976 1977 1976 1977

Fig. 3.8 A-D CC and PSc values calculated between adjacent sampling dates (and plotted midway between them) for invertebrate communities at each site group. Unlinked symbols represent comparisons of the last 1977 samples (i.e., July or October) with the nearest months in 1976 (i.e., September and November).

dominant species (PSc values over 80%) but greater variation in taxonomic composition (CC values 52 - 63%) between years. These differences between years were due primarily to reduced species richness in October

1977 (14 taxa) compa~red with the similar period in 1976 (17 - 18 taxa) (Fig. 3. 7c) .

There was greater seasonal variation in CC and PSc in site group B (Fig. 3.8B) than in site group A. Most change in abundance of dominant species (i.e., lowest PSc values) occurred between September and November 1976 and also between December 1976 and January 1977. As with site group A, CC was lower than PSc for most of the year (and low in real terms, i.e., few species) suggesting that there was considerable variation in the taxonomic composition of the community. This variation was reflected also in species richness (Figs. 3.7b and c). Although the invertebrate community at site group B showed the most seasonal change not only in community composition (as indicated by CC and PSc

values, Fig. 3.8) but also in the number of taxa present (Fig. 3.7c) I species evenness (Fig. 3.7a) and total invertebrate densities (Fig. 3.11), 69

the number of samples collected (16) was considerably fewer than the number from other site groups and the seasonal distribution of sampling effort was uneven (Table 3.16). Consequently it is difficult to separate real changes in community composition from those arising as an artefact of biased sampling effort.

The data for site group C should be the most reliable since this was the most homogeneous assemblage of sites (see Table 3.9) and was based on the largest number of samples (Table 3.16). There was very little seasonal change in species composition, with CC values between sampling dates ranging from 64.7% to 89.5% with the main peak between March and April 1977 (Fig. 3.8C).

Lower values were caused by the intermittent presence of uncommon species (e.g., leeches and some insect species). High PSc values between most sampling dates suggest that the percentage abundance of the commonest species was consistent also. The three highest PSc values (around 85%) occurred between the autumn and winter months when the communities were dominated by C. viridissima (Fig. 3.9a) and P. antipodarum (Fig. 3.9f). The low PSc between July and October 1977 was primarily due to the switch from a C. viridissima (15%) and P. antipodarum (25%) dominated community to one dominated by P. antipodarum (70%) alone. Comparison of the October 1977 samples with those from September and November 1976 showed that the communities had similar abundances of dominant species (high PSc) and taxonomic compositions (high CC) implying that changes in community composition occurred in an annual cycle (Fig. 3.8C). However, since only one cycle was investigated, it is not possible to say whether this would always be so.

In the most heterogeneous site group, 0, variations in CC between sampling dates (Fig. 3.80) were, in the main, due to the intermittent presence of various insect taxa, especially in January, March and June 1977. Fluctuations in PSc, an index that is influenced markedly by dominants, were due to particularly low percentage contributions of P. antipodarum to the invertebrate community, caused, in part, by increased representation of Cladocera and Oligochaeta (January 1977) and Chaetogaster, Cladocera, Copepoda, and Insecta (May 1977) relative to neighbouring months (Figs. 3.9b, c and d). The habitats of site group o were not sampled in October 1977 and the comparison of July 1977 with September 1976 (Fig. 3.80) suggests that communities present in these months differed, especially in terms of the percentage representation 70

of dominant taxa (e.g., Oligochaeta and P. antipodarum) (Figs. 3.9b and f). The numbers of taxa present were quite similar (19 in September 1976 and 20 in July 1977) but only 62.5% of the taxa were .shared (Fig. 3. 8D) .

3.3.6 Seasonal Changes in the Abundance of Major Taxa at the Site Groups

In detailed analysis of seasonal changes in the abundance of major taxa there were often marked variations that were difficult to explain. Sampling on some dates was more difficult than others and it is possible that between-date comparisons may be subject to greater error than comparison of samples from one date. This could be due to the influences of recent weather conditions (periods of high winds,

heavy rain, etc.) compounded with ~rue seasonal changes and/or sampling variability, weather conditions at the time of sampling and the time of day of sampling (although these last two factors were standardised as far as was possible).

2 ?', INSECTA COELE:TERATA ~ I ,

I" 0.., '0 d >< ,0, I ,.! \ I \ I I \ I I \ I £t - "•• /.. ~ ~ ..' .0 "... .1"'\ 0...... 1l.,1 \ '/~" 0. y" .'+ .. 'o-~ .'. V: ". ..l~ ~ :. B". • ... :t'; 0_ p ••• *. +.. Iit-" E OI+_~~~~tf!=.;_;,...~...... _-..,-o_-_-;:o_:'~~.OL.-, ..~-:o.~ .... ,,-' 'T'-' -.--;:..'", O!+--;:...... ~·"'-· ....'T .. '-·~-,·:r·-·r· _'T'c-~~~'"I··:-_'·_··r~,r-r~_·.;·_·~..:.Jr~· ·-'T'-'~",:-' -,-t,

~ 4 ANNELIDA ACARINA en b "­ 'J ci e c 2 >- I- ({) Z W a 20 W 0 I- 4: ~ 6 W I- ffi 4 z> '2

Fig. 3,9 a - f Seasonal changes in abundance of major taxa at each site group. (September 1976 - October 1977.)

If· •••• '. site group A, site group B, -" site group C, o~ -- 0 site group D. 71

General of seasonal changes in the abundance of or taxa at each of the site groups (Fig. 3.9) will be discussed, but detailed analyses will not be attempted. Seasonal changes in total invertebrate density (numbers per gram dry weight of macrophyte) and percentage composition of invertebrate communities at each site group are shown in . 3.10 - 3.13.

(1) Site A

The invertebrate community at site group A was dominated by Mollusca (mainly P. antipodarum) (48.2 - 89.8%) in all months sampled (Fig. 3.10). Peak abundances of molluscs occurred in March 1977, when the population was dominated by small individuals, and lowest numbers were present in December and January (Fig. 3.

(") SITE GROUP A 0..- >- x I- (J) (f) +-' ;>. Z .r= W 0- 0 0 '- (.J w ro E +-' co ~ W I- ;>. 0::: '- .w "0 .> 01 Z ~ 0 c:

~ ro +-' '- 5 ·wo E c:

S I 0 I N I F I M I A I M I J 1976 ·1977

. 3.10 Seasonal in total invertebrate and percentage composition of invertebrate communities (in terms of major taxa) at site group A during the sampling period. (Key to major taxa given in 3.2.)

Crustacea were the next mos t numerous group (1.9 - 19.8% , 3.10) with highest densities in autumn and winter (Fig. 3.9c), but sometimes C. viridissima with a greater seasonal range (0.2 - 30.5%) 72 made a larger percentage contribution to the community (Fig. 3.10)" Highest densities of C. viridissima were present between January and April (Fig. 3.9a). Densities of Acarina increased from September 1976 to March 1977 (Fig. 3.ge) with dominant species being H. lemnae and P. uncata exigua (see life-history section). The percentage contribution of mites to the community at site group A ranged from 1.3% (April 1977) to 12.1% (December 1977) (Fig. 3.10). Insecta never reached high densities at site group A and seasonal changes in abundance were difficult to interpret since low numbers of individuals of ten species contributed to the pattern (Fig. 3.9d). Generally, peaks in abundance were due to Hydroptilidae (December 1976, May 1977), Chironomidae (March, May and June 1977) and H. amabilis (June 1976). Annelida were always present in low numbers at the sites of group A and never comprised greater than 5% of the fauna at any time (Fig. 3.10). They were most abundant (about 40 per gram dry weight of macrophyte) in June 1977 (Fig. 3.9b). Overall invertebrate densities at site group A were highest in March 1977 and lowest in December 1976 (Fig. 3.10).

(2) Site B

At site group B, two macrophytes (Myriophyllum propinquum and Ranunculus fluitans) were sampled but both species were collected on only four occasions (November and December 1976, March and Ap~il 1977). In September 1976 only R. fluitans was collected, and in each of the remaining four months only M. propinquum (Appendices 2.7 and 2.8). However, at most times when both Myriophyllum and Ranunculus were collected there was a high correlation between the percentage composition of invertebrate communities (by major taxa) on both plants (2 Nov. 1976 Spearman's Rank Correlation Coefficient (Rs) 0.886,

Students t 3.816, P < 0.010; 2 Dec. 1976 Rs 0.829, t = 2.960, P < 0.025; 8 Apr. 1977 Rs = 0.886, t = 3.816, P < 0.010). Only in March 1977 was there no significant correlation (Rs = 0.421, t = 0.929, NS) due to a small sample of Ranunculus on this date (Appendix 2.8). Consequently, since there was normally no significant difference between the invertebrate communities on Myriophyllum and Ranunculus (in terms of the percentage composition of major taxa), the absence of a collection of Ranunculus or Myriophyllum should not change markedly the overall pattern of community composition at site group B as shown on Fig. 3.11). 73

(V) 2 0 x SITE GROUP B

>- (JJ I- ...... >- (J) ..c Z 0.. W 0 0 '- u W - w '- I- -0 0::: W 01 > ~ z 0 c

o'- . 'ro E c

J I A I S o I I J M J o 1977

Fig. 3.11 Seasonal changes in total invertebrate density and percentage composition of invertebrate communities (in terms of major taxa) at site group B during the sampling period. (Key to major taxa given in Fig. 3.2.)

Usually the invertebrate community at site group B was dominated by Mollusca (35.3-85.9%). Only in September 1976 was the percentage representation of Mollusca exceeded by that of another group, viz., Crustacea (38.2%, Fig. 3.11). The pattern of molluscan abundance at site group B showed certain similarities to that at site group A although peak densities occurred one month later (in December 1976 and April 1977). The decline in abundance of Mollusca (Fig. 3.9f), and most of the other taxa, after April 1977 at site group B was due to die-back of Myriophyllum and Ranunculus in winter (and consequent sampling difficulties) . c. viridissima again showed marked seasonality with peak abundance in January 1977 (Fig. 3.9a), and absence in September 1976 and October 1977. Chlorohydra comprised between 0 and 35.4% of the invertebrate community of this site group (Fig. 3.11). Abundance of Annelida was similar to that at site group A with relatively low numbers present at all times of year (Fig. 3.9b), although at this site this group comprised up to 10.8% of the community in May and October 74

1977 (Fig. 3.11). Densities of Crustacea at site group B were, on average, lower than at any of the other site groups (Fig. 3.9c), but the seasonal pattern was much the same as at site group C if allowance is made for the absence of samples for July 1977. Peaks in abundance of Crustacea occurred in January and April 1977. Insecta comprised between 3.9% and 22.6% of the community at site group B (Fig. 3.11) with peak abundances in December 1976 and April 1977. The major contributors to the December peak were Hydroptilidae (86.7%) and of these 90.8% were early instars (i.e., pre-cased pre-fifth instar larvae). The April peak was due to pre-cased Hydroptilidae and also Chironomidae larvae. Acarina at site group B showed a clear seasonal pattern of abundance, not unlike that at site groups A and C, with a summer peak and a winter low (Fig. 3.9c). H. lemnae was the dominant mite up to and including January 1977 and thereafter Piona comprised a greater proportion of the declining population. Overall invertebrate densities at site group B generally were highest in summer and lowest in winter (Fig. 3.11). Very large numbers of small P. antipodarum were responsible for the greater part of the peak in invertebrate abundance in April 1977 (Figs. 3.9f and 3.11).

(3) Site group C

Mollusca were the dominant group in eight of the ten months sampled (Fig. 3.12) although the dominance was not as marked as at the other three site groups (Figs. 3.10, 3.11 and 3.13). Between 30.4% and 72.3% of the community at site group C comprised Mollusca, with the highest densities of P. antipodarum and G. corinna in April and May. Seasonal fluctuations in mollusc abundance in this, the most stable habitat type, were less marked and densities were generally lower than those at other site groups (Fig. 3.9f). Chlorohydra and Crustacea were almost equally numerous members of the community at site group C. Seasonality in abundance of Chlorohydra was very marked with peak numbers in April 1977; numbers in September and October were low. Chlorohydra comprised between 2.7% and 34.5%, by numbers, of the community at this site group (Fig. 3.12). There were also marked changes in abundance of Crustacea with peaks in January (Copepoda 53%, Cladocera 46%), April (Copepoda 86%) and July (Cladocera 84%) 1977 (Fig. 3.9c). Copepoda (Eucyclops serrulatus) were more abundant than Cladocera (mainly Bosmina and Graptoleberis) in all months sampled except September 1976 and July 1977. Ostracoda were of minor importance numerically in the habitats of this site group. Annelida were more 75

("V) ~ SITE GROUP C x >- I- if) Z ..c: W a. 0 0 I- U W '0 «l- E a:: - co ~ W >. l- I- a:: "0 w OJ > ': z 0 c

co x co

I- 5 0 ';0 E c

0~

F I M I J I J A M J A S o o I 1977

Fig. 3.12 Seasonal changes in total invertebrate density and percentage composition of invertebrate communities (in terms of major taxa) at site group C during the sampling period. (Key to major taxa given in Fig. 3.2.)

common in invertebrate communities at site groups C and D than in groups A and B (which were in the eastern sampling zone). At site group C, Annelida c9mprised between 2.4% and 17.1% of invertebrate numbers with peak densities over the 1976 - 1977 summer, in May 1977 and July 1977. Unidentified Oligochaeta contributed most to the pattern of abundance but the May peak was boosted by high numbers of Chaetogaster (Fig. 3.9b). Acarina abundance at site group C showed a clear seasonal pattern with a summer peak and winter low (Fig. 3.ge). This same trend was seen also at the other site groups. As at site group B, H. lemnae was dominant early in the sampling period (up to January 1977) and Piona thereafter. T. novus was present in much higher densities in January and March 1977 than at other times (see Appendix 2). The representation of Acarina in the community at site group C varied from 0.5% (July 1977) to 9.0% (November 1976). The Elodea monoculture that comprised site group C was generally the poorest macrophyte habitat in 76

Lake Grasmere for Insecta. Insecta comprised between 0.8% and 4.4% of the community at this site group and the peaks in seasonal abundance were almost entirely due to Chironomidae larvae (May and July 1977, Fig. 3. 9d) .

Overall invertebrate densities at site group C were lowest in September with peaks in January (Crustacea, Acarina), April/May (Chlorohydra, Crustacea), and July (Crustacea) (Fig. 3.12).

(4) Site group D

Mollusc densities at site group D showed similar patterns of seasonal abundance to those at the other site groups. The peak in abundance in April 1977 was in common with site group B (Fig. 3.9f). Mollusca comprised between 47.4% and 87.1% of the community at site group D (Fig. 3.13). Crustacea were the next most common group (5.3 - 31.1%, Fig. 3.13) with peak abundances in late summer and autumn­ early winter. Ostracoda were common at all sites of site group D (and at site group A but not at other site groups) and contributed most to crustacean abundance in November 1976 and May 1977. In January 1977 Cladocera (84%) dominated the Crustacea, whereas Copepoda were dominant in March (57%) and June (64%). In May, the percentages of Copepoda and Cladocera were about equal (39 - 40%) with Ostracoda (20%) comprising the balance of the Crustacea. Site group D encompassed the most frequented habitats for insects in Lake Grasmere, both in terms of numbers of taxa (14 species plus about nine species of Chironomidae) and numerical abundance (Fig. 3.9d). Insecta constituted between 3.3% and 18.4% of the fauna at this site group (Fig. 3.13) and showed marked seasonality. Peaks in abundance, mainly paroxyethira hendersoni and Chironomidae, occurred in January and May - July 1977 (Fig. 3.9d). Annelida were also most represented in site group D (0.6 - 25.3%, Fig. 3.13) relative to the other site groups. The pattern of abundance was similar to that at site group Cexcept that peaks were more pronounced. As at site group C, the May peak was boosted by large numbers of Chaetogaster. The July peak was mostly unidentified oligochaetes (84%) and the January peak comprised about 57% unidentified oligochaetes and 43% Chaetogaster. Site group D comprised sites in which mites, especially the oribatids H. lemnae and T. novus, were well represented. The early dominance of H. lemnae (to January 1977) was a feature shared with the other site groups, however the dominance of Piona was short-lived (only March 1977). 77

C0 GROUP 0 C) ..-- x > ID r- ~ U) ~- z CL w 0 ~ 0 0 w ~ r- E 4 ~ -~ m ~ w ~ r- ~ ~ W ~ > ~ Z 0 c

x~ ~

~ 5 0 'ro E c

0~

S I 0 I N o J F I M I A I M I J. J A I S I 0 I 1975 1977

Fig. 3.13 Seasonal changes in total invertebrate density and percentage composition of invertebrate communities (in terms of major taxa) at site group D during the sampling period. (Key to major taxa given in Fig. 3.2.)

H. lemnae (April and October 1977) and T. novus (May, June and July 1977) were dominant thereafter. Larval T. novus were most abundant in late summer and autumn (see Appendix 2). Mites comprised between

1.7% and 11~0% of the fauna at site group D (Fig. 3.13). Chlorohydra was not well represented at this site group (0- 3.7%) (Fig. 3.13). However the seasonal pattern of abundance was similar to that at site group B with maximum numbers in January 1977.

Total invertebrate abundance at site group D was usually greater than at any other site group but the seasonal pattern was most similar to that at site group A and was influenced to the greatest extent by mollusc abundance (mainly P. antipodarum).

3.4 CONCLUSIONS

3.4.1 The Invertebrate at the Site ------~~------~--~--~~~~~~ in Terms Taxa 78

The most notable feature of the taxonomic composition of invertebrate communities of the macrophyte zones of Lake Grasmere was the numerical dominance exhibited by a relatively small number of species. The 13 taxa of species group 6 comprised between 95.5% (site group 0) and 99.7% (site group C) by numbers of the invertebrate communities at the site groups (Table 3.17). The dominance of species group 6 (which comprised those species that occurred at every site sampled) was, to a large extent, due to P. antipodarum which was numerically dominant at each site group (36.2 -70.6%) (Table 3.18). Only at site group C was its dominance approached by C. viridissima (24.9%). Furthermore, only six taxa (P. antipodarum, C. viridissima, E. serrulatus, G. corinna, Cladocera, and H. lemnae) contributed more than 5% (by numbers) to the fauna at any site group. The numerical contribution of the remaining nine species groups to invertebrate abundance was minor (0.3 -4.5%) (Table 3.18). These minor species groups (i.e., all except species group 6) contributed to varying degrees to the communities of the four site groups (Table 3.17). The benthic species of group 7 were next in abundance to species group 6 at site groups 0 (3.8%), A (1.2%) and C (0.2%) whereas at site group B species

group 5 (= P. tillyardi) filled second place (2.01%). No other species group contributed more than 0.75% to the invertebrate community at any site group (Table 3.17). Marked differences were evident in the contributions of three species groups to the communities at groups A and B on the one hand, and C and 0 on the other. Species groups 1 and 5 were more prominent members of the communities at eastern sites (A and B) than in other areas, whereas X. zealandica (species group 8) was found only in site groups C and 0 (Table 3.17). Although Mollusca were the dominant taxonomic group (by numbers) in the invertebrate communities at each site group (43.7 - 72.6%) (Table 3.19) and P. antipodarum was the most common species at all sites, there were marked differences in the distributions of G. corinna, P. variabilis and S. novaezelandiae (Tables 3.9 and 3.20).

Crustacea and Coelenterata were the only other major taxa that constituted greater than 10% of the fauna of any site group (Table 3.19). C. viridissima (Coelenterata) was a prominent component of the community at site group C (primarily due to its seasonal appearance at very high densities), whereas Crustacea constituted greater than 10% of the communities at site groups C (due to planktonic immigrants), 0 and A. Annelida had proportionately greater representation at sites with benthic influences (especiallY at site group 0). Mites (except for 79 site group B) and insects (except for site groups B and D) were less than 5% of the fauna at any site group.

Table 3.17 Percentage contribution (by numbers) of the individual taxa of the different species groups to invertebrate communities of the site groups (September 1976 - October 1977). *:: <0.01%, - '" not present. (See also Table 3.18 for details of species group 6.)

Site group Taxa A B C D

group 1 0.25 0.67 0.01 0.04 Sigara arguta 0.07 0.01 * Hudsonema amabilis 0.21 0.56 * 0.04 variabilis 0.04 0.04 group 2 0.09 0.05 0.04 Glossiphonia multistriata 0.02 * * Diaprepocoris zealandiae 0.07 0.04 0.03 group 3 0.08 0.41 0.15 pycnocentrodes aureola 0.04 0.06 Nymphula nitens 0.04 0.41 0.09 group 4 0.04 0.02 0.17 (Triplectides cephalotes) group 5 0.73 2.01 0.05 0.11 (Paroxyethira tillyardi) group 6 97.61 96.61 99.66 95.47 (see Table 3. 18) group 7 1. 20 0.26 0.17 3.76 Nematoda 0.12 0.07 0.12 0.87 Ostracoda 0.71 0.15 0.03 2.31 um novaezelandiae 0.37 0.04 0.02 0.58 group 8 0.04 0.05 (Xanthocnemis zealandica) Species group 9 * * 0.17 Oxyethira albiceps 0.08 Oecetis unicolor 0.08 Arrenurus sp. * * 0.01 group 10 0.04 0.03 Antiporus strigosulus 0.01 Deleatidium sp. * 0.01 Zelandobius furcillatus 0.03 0.02

The ranking of major taxa at each site group on a percentage by numbers basis (e.g., at site group A : Mollusca 1, Crustacea 2,

Coelenterata ~ 3, Acarina 4, Insecta ~ 5 and Annelida ~ 6, Table 3.19) tends to bias representation in favour of taxa with high densities that 80 may be highly seasonal (e.g., Coelenterata at site group C). A more realistic assessment of the year-round ranking of taxa in the communities at site groups may, perhaps, be obtained by using a points method that gives equal weight to each month's samples from each site group (see Sanders 1960). Table 3.21 lists the biological index values of the major taxa determined first by ranking the taxa according to abundance from 1 to 6 for each month for each site group. A rank of 1 is then given a value of ten points, a rank of 2 nine points, and so on, and the points added to give the biological index value. Thus, if a taxon is ranked first in each of, say, ten months it will gain 100 points, the highest possible score. The highest possible biological index value for taxa in site groups A, Band D was 90 (nine months' data) and for site group C 100 (ten months' data).

Table 3.18 Percentage occurrence (by numbers) of invertebrate taxa comprising species group 6 at each site group. Data from all quantitative samples combined (September 1976 - October 1977) .

Site group Abundance Taxa ranking A B C D

Potamopyrgus antipodarum 1 70.56 69.47 36.17 53.88 Chlorohydra viridissima 2 8.11 6.52 24.94 1.31 Eucyclops serrulatus 3 8.37 5.06 12.44 7.56 Gyraulus corinna 4 1. 59 0.82 7.53 6.99 CLADOCERA 5 1.44 2.16 7.67 6.50 OLIGOCHAETA (except 6 1.49 0.15 4.01 4.12 Chaetogaster sp.) Chaetogaster sp. 7 0.25 1.15 2.07 2.87

Hydrozetes lemnae 8 1. 90 5.06 1.03 2.49

CHIRONOMIDAE 9 0.79 1. 38 1. 08 4.09 Piona uncata exigua 10 L81 2.12 1.52 0.56 Paroxyethira hendersoni 11 0.50 2.12 0.68 3.24 Trimalaconthrus novus 13 0.49 0.34 0.41 1.45 Cura pinguis 16 0.31 0.26 0.12 0.42 All other taxa (see 2.39 3.39 0.33 4.52 Table 3.17) 81

Table 3.19 contribution of the major taxa (% by numbers) to the composition of invertebrate communities at the site groups, and for all site groups combined. (Data from all quantitative samples combined, September 1976 - October 1977. )

Site group Taxa Overall A B c D

Coelenterata 8.11 6.52 24.94 1. 31 14.75 Annelida 2.19 1. 63 6.36 8.29 5.57 crustacea 10.52 7.37 20.14 16.37 16.40 Insecta 2 ..42 6.59 1. 92 8.07 3.61 Acarina 4.20 7.52 2.96 4.51 3.80 Mollusca 72.56 70.37 43.72 61.45 55.87

Table 3.20 Percentage composition of Mollusca at each site group (all quantitative samples combined, September 1976 October 1977).

site group Taxa A B C D

Gyraulus corinna 2.21 1.16 17.29 11.38 Physastra variabilis .0.07 0.05 potamopyrgus antipodarum 97.21 98.73 82.66 87.68 Sphaerium novaezelandiae 0.51 0.05 0.05 0.94

Ranking of major taxa within site groups using biological index values (Table 3.21) gave a slightly different indication of their numerical importance at the site groups compared with the overall percentage composition of taxa (Table 3.19). The differences were most marked at site group B where Mollusca were the only group to retain their (first) ranking. The most notable difference at site group B was in the ranking of mites which were second on an overall percentage basis (Table 3.19) but only fourth on a points basis (Table 3~21). This change was due to marked seasonal variation in mite abundance at this site group (Fig. 3.18). 82

At the other three site groups the differences between the percentage and points methods were less marked and, except at site group 0, involved single changes of only one rank. At site group D the extreme seasonality of Annelida (F.ig. 3.9b) resulted in a higher ranking on a percentage basis (Tables 3.19 and 3.21). Although most differences appear relatively minor I believe that the biological index values give a more realistic overall assessment of the numerical importance of each of the major taxa at each site group. This method of evaluation is not influenced markedly by seasonal 'explosions' of abundance, except in so far as they affect ranking within the particular month/s of great abundance.

Table 3.21 Biological index values for major taxa at each of the four site groups (September 1976 - October 1977). Values for

site groups A, Band 0 based on 9 months' data; site group C, 10 months' data.

site Rank Biological Taxon group 1 2 3 4 5 6 index value

A Mollusca 9 90 Crustacea 5 2 2 75 Acarina 1 4 2 2 67 Coelenterata 3 2 1 3 64 Insecta 5 3 1 58 Annelida 1 1 2 5 52

B Mollusca 8 1 89 Crustacea 1 2 3 '1 2 71 Insecta 2 3 3 1 69 Acarina 3 2 1 1 2 66 Coelenterata 1 1 2 5 61 Annelida 1 1 1 2 4 56

C Mollusca 8 2 98 Crustacea 2 2 6 86 Coelenterata 4 2 1 2 1 76 Annelida 1 1 6 2 71 Acarina 1 1 2 3 3 64 Insecta 1 3 6 55

0 Mollusca 9 90 Crustacea 5 3 1 75 Insecta 1 3 4 1 68 Acarina 1 2 1 4 1 61 Annelida 2 1 3 2 59 Coelenterata 1 8 46 83

3.4.2 and Invertebrate Different Site

Seasonal changes in species diversity, species evenness, species richness, and invertebrate densities have been discussed in previous sections. The following discussion is concerned with the differences or similarities in overall community composition at the site groups (all data combined, 2 september 1976 - 3 October 1977) .

Overall species diversity was highest at site groups e and D with diversity at site group e having a greater evenness component than at site group D (Table 3.22). The E. canadensis 'monoculture' of site group e was the most homogeneous macrophytic habitat in Lake Grasmere and the one least subject to seasonal change (e.g., Fig. 3.8e). These factors tend to enhance species evenness.

The communities at site group D had 32 taxa, and the highest species richness component of diversity, whereas only 26 taxa were collected from site group e and this, when related to sample sizes, resulted in the relatively low value for Whittaker's species richness for site group e (Table 3.22). Habitat heterogeneity, which makes available a greater number of niches, was considered to be the main cause of high diversity (especially species richness) at site group D.

Table 3.22 Overall species diversity (H'), species evenness (J'), species richness (d), numbers of taxa, mean invertebrate densities (numbers per sample and numbers per g dry wt of macrophyte) and numbers of samples for each of the four site groups (September 1976 - October 1977).

Site group A B e D

H' 0.54 0.57 0.80 0.80 J' 0.38 0.43 0.56 0.53 d 6.43 6.42 5.80 7.69 No. of taxa 27 22 26 32 Mean invertebrate density: No./sample 334.83 167.30 392.99 369.90 No./g 690.83 446.75 651.36 1064.65 No. of samples 47 16 78 39 84

The eastern sites (site groups A and B) were different, with lower total diversity and species evenness (Table 3.22). Whittaker's species richness values were similar for both of these si t,e groups due to the relatively lower number of taxa collected from site group B being compensated for by (or perhaps related to) lower numbers of invertebrates in samples or fewer samples collected from site group B (Table 3.22). Species groups 2, 8, 9 and 10 were not collected in samples from site group B (Table 3.17) during the quantitative sampling program. Certain species (especially in species groups 2 and 10) possibly may have been present since they occurred in the adjacent sites of site group A (Table 3.17). It is likely, therefore, that the similar species richness values, as indicated by Whittaker's index, are a true indication of similarities between invertebrate communities at site groups A and B (Table 3.22).

Invertebrate densities at eastern sites (site groups A and B) tended to be lower than elsewhere when expressed as numbers of invertebrates per sample. Highest densities were recorded from the most homogeneous site group (C) and lowest densities from the least intensively sampled site group B (Table 3.22). Expression of invertebrate density as numbers per g dry wt of macrophyte habitat altered the situation. Highest densities were then recorded from site group D and lowest again from site group B. Site groups dominated by E. canadensis (C and A) showed the least proportional increase in density relative to other site groups when invertebrate densities were converted from nos./sample to nos./g dry.wt ofmacrophyte (Table 3.22). The stunted macrophytes and the inclusion of benthic animals at site group D were considered to contribute to high invertebrate densities on a numbers/dry wt of macrophyte basis',

3.4.3 Community Similarity

Table 3.23 presents average values of CC and PSc between sampling dates and for replicate samples from individual sites for each site group. Replicate samples are, by definition, expected to be of similar composition since they are collected from the same area of the habitat on the same sampling date. PSc values for replicate samples were high (> 66%) suggesting that the sampling method was good for estimating the percentage abundance of the commoner species. Generally, values of CC were lower and more variable (42 - 70%) indicating that, especially in some habitats (e.g., E I and E M), there was a greater 85 proportion of relatively uncommon species and the sampling procedure was not adequate for these species.

Average values of CC and PSc between sampling dates cannot be compared directly with values from replicate samples since the time periods between sampling dates influence markedly the numerical values of CC and PSc obtained.

However, an indication of the overall amount of seasonal change in invertebrate abundance was obtained by sum of squares analysis of a 'site X times' matrix of invertebrate densities (in terms of numbers! sample) (i.e., 13 sites X 10 sampling times). The sum of squares variation in total invertebrate density at sites was 21,499.6 and over times was 43,843.5. This shows that there was 2.039 times more 'variation heterogeneity' in times than sites. In other words, temporal changes were more pronounced than spatial variations.

Table 3.23 Average values of CC and PSc (± standard deviation) between sampling dates and for replicate samples from individual sites, for each site group (September 1976 - October 1977).

cc PSc

Site group A (between dates) 69.87 ± 9.65 77 .02 ± 7.97

E E 2 (replicates) 59.40 ± 12.36 83.53 ± 8.05

E I (replicates) 42.00 ± 7.25 66.20 ± 4.55

Site group B (between dates) 52.26 ± 16.48 64.94 ± 13.30 E M (replicates) 47.00 ± 4.24 79.00 ± 2.83

Site group C (between dates) 73,08 ± 7.42 74.30 ± 12.22

N E 2 (replicates) 68.37 ± 14.77 74.73 ± 13.13 WE 2 (replicates) 69.97 11. 58 69.07 ± 12.01

Site group D (between dates) 71.52 ± 11. 77 70.45 ± 12.25 S I (replicates) 65.75 ± 14.01 74.25 ± 11.38

W I (duplicate) 55.60 71.63

87

CHAPTER IV

TROPHIC INTERRELATIONSHIPS 'OF SOME MACROPHYTE-ASSOCIATED INVERTEBRATES

4.1 INTRODUCTION

In recent years it has become increasingly apparent that knowledge of trophic relationships is a necessary prerequisite to an understanding of energy flow and community dynamics in freshwater ecosystems (e.g., Cummins 1973). Compared with rocky shore, stream and sediment habitats, the macrophyte beds of a lake present, to the animals that live in them, a habitat of much greater physical and chemical complexity. Because of this, and formidable problems of sampling (see Chapter III), it is not surprising that the investigation of macrophyte-zone animals of lakes has progressed little beyond the stage of observation and lags behind studies of stream animals (Moss 1980).

Very complex food webs must exist within macrophyte beds (Moss 1980) .' The study of trophic relationships in freshwater communities reveals a great diversity of available foods and a variety of methods of resource utilisation (Anderson & Cummins 1979). Because of such complexity, attempts have been made to elucidate nutritional relationships of aquatic invertebrates by defining trophic categories in terms of the food itself (herbivores, detritivores, carnivores, i.e., trophic levels) or according to the way it is obtained (e.g., the shredder, collector, scraper and.predator functional groups of Cummins (1973». Almost all stream invertebrates are omnivores but usually take a proportion of plant or animal food at particular of their life histories (Moss 1980). Dietary shifts in later instars are well documented for a number of species (Winterbourn 1971b, Crosby 1975, Fuller & Stewart 1977, Anderson & Cummins 1979). In these instances difficulties are encountered in assigning invertebrates to fupctional groups. Nevertheless, the concept of functional groups has proven useful in analysing the partitioning of food resources on a community basis in lotic ecosystems (Cummins 1973, 1974, Merritt & Cummins 1978, Mackay & Wiggins 1979, Wiggins & Mackay 1978, Cowie 1980).

The food habits of freshwater invertebrates in New Zealand are relatively little known. Although recent works by winterbourn (1974), 88

Crosby (1975), Devonport & Winterbourn (1976), Winterbourn & Davis (1976), winterbourn (1978) and Cowie (1980) have documented the specific food habits of a number of stream insects (mainly Megaloptera, Ephemeroptera, Plecoptera and Trichoptera), very little work has been done on invertebrates living in lentic environments. Babington (1967) examined the gut contents of three species of leptocerid (Trichoptera) larvae but not in detail, and Greig (1976) investigated the feeding of a species of the more typically lotic Deleatidium (Ephemeroptera) in Lake Grasmere. Most information on the diets of lake-dwelling invertebrates in New Zealand is non-quantitative and based solely upon observation or generalisations from work on related overseas species. Consequently, a primary aim of this study was to obtain quantitative information on the food habits of a number of lake­ dwelling invertebrates in order to gain some insight into the organisation of food webs within the macrophyte zone.

The choice of study animals was influenced by several considerations, including that they had to be present in large enough numbers to make regular sampling practicable. I wanted to examine the food habits of a range of invertebrates that could be expected to belong to different functional groups (e.g., a shredder, Nymphula nitens (Lepidoptera) and a predator, xanthocnemis zealandica (Odonata». Also, it was anticipated that a study of closely related species or genera (e.g., the two hydroptilid caddisflies, Paroxyethira hendersoni and P. tillyardi, and the two leptocerid caddisflies, Hudsonema amabilis and Triplectides cephalotes) and more detailed examination of species that had been investigated previously in other lakes and ponds (e.g., T. cephalotes and H. amabilis (Babington 1967), X. zealandica (Crumpton 1979) would provide interesting comparative data. A further logical choice, in view of its numerical dominance in the macrophyte beds of Lake Grasmere, and abundance in other freshwater habitats, was the snail, Potamopyrgus antipodarum.

The food habits of freshwater invertebrates have been studied using a variety of methods, viz., field observations (Welch 1924, Berg 1941, Bay 1972), laboratory observations (Welch 1916, 1924, Berg 1941, Nielsen 1948, Babington 1967, Satija 1974), food preference experiments (Lawton 1970a,Skoog 1978, Anderson & Cummins 1979 and references therein, Anderson & Sedell 1979 and references therein), gut content analyses (Muttkowski & Smith 1929, Slack 1936, Nielsen 1948, Hanna 1957, Mecom & Cummins 1964, Babington 1967, Tarwid 1969, Thut 1969, Winterbourn 1971a, 1971b, 1973, 1974, 1978 , Satija 1974, Crosby 1975, Soszka 1975b, Devonport & Winterbourn 1976, Winterbourn & Davis 1976, Crumpton 1979, 89

cowie 1980), faecal analyses (Pritchard 1964, Lawton 1970a, Soszka 1975b, Skoog 1978, Thompson 1978a and b), serological techniques (Davies 1969, Davies, Wrona & Linton 1979) and radioactive tracers (Rodina & Troshin 1954, Sorokin 1966, Hargrave 1970, Sedell 1972, Nash 1974, Greig 1976).

Gut content analyses have been the usual methods used to determine the nature of food ingested by freshwater invertebrates. Resistant remains - spines, mouthparts, carapaces, diatom frustules and macrophyte cell walls - give clues to what has been , but soft­ bodied prey (e.g., flatworms (Davies 1969, Davies & Reynoldson 1969, 1971» may leave no trace. In addition, the presence of a material in the tract does not prove nutritional importance and differential rates of of food items also renders the nutritional interpretation of visual ,gut content analyses difficult (Cummins 1973). The most rapidly digested items, and therefore the least to be observed, might be of greatest nutritional significance. In of these problems, gut content analyses have proven useful in giving an indication of the trophic relationships of aquatic invertebrates.

The technique of faecal analysis was chosen in this study. It has disadvantages similar to those described above for content analyses, but is less time consuming and more easily to invertebrates that are too small to dissect easily.

Comparisons of faecal pellet analyses and gut content made early in the showed no significant qualitative differences between the two methods. Since most investigators have equated the composition of contents with food intake (i.e., ingestion), I have assumed, for in the following discussion, that the composition of the faeces agrees with the composition of food ingested.

4.2 METHODS

4.2.1 Faecal ------~~--

(1) Field The species whose faecal material was analysed quantitatively (Table 4.1) were collected from areas of the lake where they were most abundant. P. tillyardi, H. amabilis and P. antipodarum were collected from shallow beds of E. canadensis and I. alpin us in the eastern sampling area; x. zealandica, P. hendersoni and T. cephalotes from the E. canadensis 'monoculture' in the northern and western sampling areas; and Nymphula nitens from M. propinquum in the eastern sampling zone. The animals were obtained using a 200 ~m 90

Table 4.1 Size classes of invertebrates used in faecal analyses. htl = hind-tibia length, hw = head width, tsh = total shell height.

Size group Size range Instar (parameter measured) (rom)

xanthocnemis zealandica 1 , 2.0 (hw) 2 2.1 3.0

3 ~ 3.1 F

Nymphula nitens 1 ( 0.875 (hw) 2 0.876 -1.100 3 > 1.101

paroxyethira hendersoni 0.23 F (hw)

paroxyethira tillyardi 0.18 F (hw)

Hudsonema amabilis ( 0.30 2 (htl) 0.31 - 0.45 3 0.46 - 0.75 4

~ 0.76 5

Triplectides cephalotes ~ 0.50 2 (htl) 0.51 - 0.80 3 0.81 - 1.50 4

~ 1.51 5

potamopyrgus antipodarum * ~ 3.50 (tsh)

* P. antipodarum of a smaller size class (tsh < 3.50 rom) were collected also but densities of food material on filters were not enough to make enumeration worthwhile. The division between the size classes represents the size a"t which P. antipodarum become mature (Winterbourn 1970b) •

mesh hand-net on the end of a 2 m aluminium pole. The net was moved to and fro through the appropriate submerged macrophyte beds and its contents were emptied into a white tray for sorting. All individuals of the chosen species were removed from the sample in the tray and isolated individually in 40 rom x 25 rom snap-top glass tube vials containing about 3 ml of 91 filtered tap water. This procedure was continued until at least ten individuals of the dominant size classes had been collected. The remainder of each sample was transferred to a bag and preserved with formalin for subsequent microscopic and use in life-history analyses (Chapter V) .

The species under study required different treatments to obtain best results. All were washed in filtered water prior to isolation and the cases of T. cephalotes larvae were removed to minimise contamination by non-faeces-derived material, and so that the starved larvae could not feed on their cases. Hydroptilidae, because of their minute size, yielded best results when several individuals were placed in each vial.

(2) Laboratory procedure. The isolated invertebrates were left to produce faeces over a period of 12 to 24 hours. The animals were then removed from the vials, measured, and in 70% alcohol. About 1 ml of absolute alcohol was added to each of the vials to prevent bacterial or fungal interference occurring prior to further treatment. The lid of each vial was labelled with the size of the invertebrate from which the faeces had been obtained.

The contents of each vial were transferred into a plastic 15 x 50 rom vial, using filtered water as necessary, and subjected to 10- 30 seconds of ultrasonic vibration in a Bransonic 3 ultrasonic generator half-filled with water. This broke up the faecal material and dispersed it evenly throughout the liquid. Plastic vials withstood the rigours of repeated ultrasound better than glass. The suspended faecal material was washed into a small (16 rom internal diameter)

Millipore funnel and filter column fitted with a 0.45 ~m pore size filter paper. The filter column was fitted with a perspex 'reduction disc' of my own design to reduce the filtering area to 10 rom diameter. The aim of this was to increase the density of faecal material on the filter and so facilitate enumeration, and to keep individuals' faeces separate as far as

The filter paper used was Gelman Metricel GA-6 size 0.45 vm) which was obtained in 200 x 250 rom sheets (Part No. 60180). Squares of 12 x 12 rom (approx.) were cut, ready for use, with a scalpel. (This method was more economical than that ready-made 25 rom diameter filters (Part No. 60172) and also enabled up to four filters to be mounted on each microscope slide, thus saving on materials, storage space and handling time) . After filtration, under low vacuum, 92

the filter papers were allowed to dry, under cover, at room temperature. They were then mounted on standard microscope slides using lactophenol­ PVA stained with Lignin Pink. After about a week in a slide-drying oven (at 66°C) the slides could be examined.

Slide-mounted filters were examined at 450x magnification under phase contrast using a Meopta C36 Bi Microscope (Type 56243). Six fields of view were examined per filter. Diatom genera or species were recognised and counted and the projected areas of macrophyte, detritus, filamentous algae, and animal remains were estimated using a gridded eyepiece. The counts of each diatom genus or species were converted to an areal basis by mUltiplying by standard areas (Table 4.2) calculated from measurements of several diatoms of the same type and size. Patrick (1959), Barber (1962), and Weber (1971) were useful aids to identification, as were a preliminary list of diatom genera from stony substrates in Lake Grasmere given by Greig (1976) and the list of planktonic diatoms recorded by stout (1972).

In order to facilitate the identification of animal remains in faecal pellets, a reference collection of potential prey invertebrates was prepared (using hot KOH, as necessarYt to digest away soft tissues). Some of the criteria by which prey items were recognised are shown in Fig. 4.1. In addition, Nematoda were occasionally present intact in faecal pellets; the cladoceran, Simocephalus vetulus was recognised by the characteristic postabdomen and carapace; the cyclopoid copepod Eucyclops serrulatus was represented by typical limb segments and furcal rami; and ostracods were recorded as prey by the presence of valves and limbs. Among the insects recorded as prey items, hydroptilid caddisfly larvae were identified by the species-specific shapes of the frons, tarsal claws, and prothoracic sternites (see Chapter VI). Mandibles and limb segments were useful also but did not enable specific identification. Chironomid larvae were also readily identifiable, using head capsules, which often remained intact, or mandibles, labial plates and posterior proleg claws (see Chapter VI). In addition to the distinctive radula teeth, the gastropod mollusc potamopyrgus antipodarum was recognised by the presence of shell fragments.

The sizes of P. antipodarum ingested were determined by measuring the total length (1) of an inner marginal tooth of the radula (Fig.4.1J). All inner marginal teeth in the radula of snails of a given size were the same size. The following relationship was found between total shell height of P. antipodarum and the length of the 93

Table 4.2 Standard areas used to calculate the projected areas of diatoms in the faecal analyses of freshwater invertebrates from Lake Grasmere. Specific names of diatoms should be regarded as tentative ly those species previously unrecorded from the Cass area). previously recorded general species are denoted by the following superscripts: S = phytoplankton, Lake Grasmere (Stout 1972) G periphyton on stones, Lake Grasmere (Greig 1976) B streams at Cass (Barber 1962) D Middle Bush Stream, Cass & Winterbourn 1977) C Cass area (checklist in Burrows 1977).

2 4 Diatom Genus ected area (rom x 10- ) (specific name/s)

S C AsterionellaG 3.10 '( formosa) , CocconeisG,D 2.18 (placentula)B,C CyclotellaG 1.77 (kuetzingiana)S,C CymbellaG,D 16.30 (cistula)C Diatoma 2.70 (elongatum) S, B,C 10.50 (hiemale)C EpithemiaG,D 5.3 (sorex)C 12.20 (turgida)B,C Eunotia 8.00 (cassiae)B Fragi 1. 68 (construens)B,C 3.05 (leptostaurum) GomphonemaG,D 5.00 (constrictum)C 10.50 (herculeanayC 0.44 (olivaceum) 2.02 (granulata)S,B,C NaviculaG 6.80 (radiosa)B,C 4.73 (reinhardii) Opephora 20.0 (martyi)B,C pinnularia 28.70 (gibba)B,C RhoicospheniaD 3.24 (curvatit) RhopalodiaG,C 1. 70 (sp.) stephanodiscus 2.85 (dubius) synedraG 1.63 ( actinastroides) 5.75 (acus)C Tabellaria 1.00 (flocculosa)B,C

inner marginal tooth by dissecting radulae from snails of various sizes (2.6 rom to 8.6 rom tsh), mounting them on microscope slides, and measuring a number of teeth.

P. antipodarum tsh = 94.013 (length of inner marginal tooth) - 3.224 rom (r=0.998). Therefore, this equation was used to determine the sizes of P. antipodarum ingested by H. amabilis and X. zealandica larvae. 94

Fig. 4.1 Skeletal remains used to recognis'e prey items in faecal analyses. (Scale bars = 0.05 mm) Acarina a. piona uncata exigua: part of genital plate of male. b. Trimalaconothrus novus: tarsus and tarsal claws of hind limb. c. Hydrozetes lemnae: tarsus and tarsal claws of hind limb. d. piona uncata exigua: mandibular claw. Oligochaeta e. Lumbriculus variegatus: seta. Cladocera f. Graptoleberis testudinaria: postero-lateral margin of valve. g. Alona sp. postabdominal claws. Mollusca h. i. j. Potamopyrgus antipodarum: radula teeth (lateral, central and inner marginal) (showing length measurement, see pp. 92, 101-103 & 114-115). Chironomidae k. Calopsectra funebris: hook from posterior proleg. 95

The number of prey items present in each pellet was estimated easily for most prey s from the number of mandibles, head plates, etc. Oligochaetes and P. antipodarum, however, were recorded only as single occurrences whenever their characteristic remains were found in each pellet.

Most of the major food categories recognised during this study are well known (viz., diatoms, detritus, macrophyte, filamentous algae and animal remains) but the final category remains a mystery, S.R.T'S, ('small round things') denoted by the diagonal striping on Fig. 4.2

(and subsequent figures), were about 9 ~m in diameter and composed of a 'solid' core surrounded by a 'membrane' with a narrow space between. They were not like eggs of invertebrates, but rather were most likely to be of or fungal origin (pollen or spores?). They were never important quantitatively.

The importance of careful preparation of permanent slide-mounts of faecal material during studies of food habits should not be under­ estimated because ability to recognise various food items (especially fragments of prey) improves with experience. It is extremely valuable to be able to refer back to 'problem' slides which, in many cases, can be successfully re-analysed.

4.2.2 Periphyton Analysis

Lengths of macrophyte stem (E. canadensis and M. propinquum) were collected from various sites. in each of the sampling areas of Lake Grasmere. The stems were stored in 70% alcohol prior to further treatment.

Lengths of 15 rnrn were cut in series down each stem beginning at the shoot tip. The periphytic coating of each length was removed by repeated ultrasonic vibration and washing in filtered water. Five cycles of ultrasonic treatment (30 seconds each) and washing with filtered water were sufficient to remove practically all of the periphyton from the plant surfaces. This was confirmed by microscopic examination of the plant surfaces and by examination of the filtrate from a sixth washing, which showed that very few diatoms remained attached to the plant. The washings from each section of stem were filtered through a Millipore filter column, and each filter paper was mounted on a microscope slide. Slide-mounted filters were examined as for invertebrate faeces (p. 92). 96

4.3 RESULTS AND DISCUSSION

4.3.1 Introduction

The results of faecal analyse~ are presented together with preliminary investigations on the nature of the periphytic coating on macrophytes, as a potential source of food. A comparison of the faecal analyses of the seven species examined (all months and ins tars or size classes combined) is given in Fig. 4.2. Where appropriate"differences in food habits between size classes or instars of the study species are discussed later, and the findings compared with previous work on the same, or related species from New Zealand or overseas.

.!:::rol'_ c: .c: 0 - III ~I:x j~0'"0 I- c: ;/ ~i~.~ Q.I ZlC: 10 .c: CLI

III IIIro ..0 ~co Q.I ....ro

Q.I u c: co 50 '"0 c: crophyte :J ..0ro Q.I > ..-ro Q.I '- ~ 0 0 0 0 0 N= 73 74 145 64 43 126 48

Fig. 4.2 Composition of faecal material of seven invertebrate species from Lake Grasmere, in terms of major food categories, on a projected area basis (July 1976 - November 1977). (All size classes combined, N = number of individuals examined,) Also given is the key to food categories (as used on this, and subsequent figures).

Very few obvious seasonal changes were apparent in either the composition of the faecal material in terms of major food categories (i.e., diatoms, detritus, macrophyte tissue, filamentous algae, animal remains and S.R.T's), or the generic composition of the diatom category. 97

4.3.2 Periphyton Analyses

Although initially I had intended to compare the composition of faecal material with the composition of available food in the environment on a regular basis (from periphyton and invertebrate sampling), it soon became apparent that this was not a practical proposition. The seven species whose faecal pellets were analysed were collected from different areas of the lake (see p. 89) and, in some cases, were undoubtedly feeding in different microhabitats. The work involved in regular quantitative sampling of periphyton from different microhabitats and different areas of the lake proved to be prohibitive. However, a few generalisations can be drawn from the limited number of quantitative periphyton analyses made and the more extensive qualitative studies.

Pieczynska (1964) found that there was greater variability in species composition in old than young periphyton, and that various types of substrate had very similar communities provided that they were located in the same environment (e.g., in the same part of the littoral zone) . This was also the case in the present study. Semi-quantitative and qualitative examinations showed great similarity in the composition of periphyton communities on different macrophytes from the same sampling areas in Lake Grasmere. On the other hand, certain differences were evident between periphytic communities from different areas of the lake. Notable amongst these differences was the apparently greater representa­ tion of Rhoicosphenia in diatom communities in the eastern sampling area compared to the western zone. Most other differences were minor and would add only confusion if attempts were made to correlate them with, for example, the food preferences of periphyton-feeding invertebrates.

Table 4.3 gives a typical example of the change in composition of the periphyton community in a 'transect' down a stem of E. canadensis. There was a dramatic increase in the quantity of periphyton with age of the plant stem, as measured microscopically in terms of relative densities of periphytic organisms on membrane filters (Table 4.3 (a». Diatoms comprised the greatest percentage of the community and there was a general trend to increased abundance and species richness with age down the plant stem reflecting the duration of stem exposure to potential colonisation. Although detritus made a greater percentage contribution to the periphyton community of the shoot , the quantities of detritus on older parts of the stem (in absolute terms) were higher (up to 2.5 times those on the shoot tip, Table 4.3 (a». 98

Table 4.3 Composition of periphyton communities on successive 15 mm lengths of Elodea canadensis stem collected from the eastern sampling area of Lake Grasmere (8 April 1977). (a) Percentage composition and relative amounts of periphyton. (b) Numerical composition (%) of the diatom category and total numbers of diatoms.

(a) Distances down E. canadensis stem Shoot tip 15 - 30 mm 30 - 45 mm 45 - 60 mm 60 75 mm

Diatoms 83.4 95.0 94.6 98.3 97.0 Detritus 16.6 5.0 5.4 l.l 2.0 Filamentous algae 0.6 l.0

Relative peri- phyton density 1 7.6 7.7 19.5 20.8 (by projected area)

(b) Cocconeis 97.2 94.8 95.7 98.4 98.7 Epithemia 2.8 l.0 0.7 0.2 0.3 Diatoma 2.6 2.0 0.1 0.5 Synedra 0.3 0.7 0.2 Navicula 0.7 0.2 0.2 Rhoicosphenia l.0 0.3 0.1 Gomphonema 0.3 0.3 0.4 Cyclotella 0.3

Total numbers 36 304 302 842 876 of diatoms

4.3.3 Invertebrate Faecal

(1) hendersoni and P. till Hydroptilidae) • More than 90% of the faecal material of final ins tar larvae of the two species of Paroxyethira was browsed from plant surfaces and the remainder was macrophyte tissue itself (Fig. 4.2). Diatoms predominated over other food categories in most months (Appendices 6.la and 6.2a) and overall (Fig. 4.2). Detritus and filamentous algae were almost equally represented in the faeces of both species, with the former perhaps marginally more abundant in faeces of P. tillyardi (Appendices 6.la and 6.2a). Macrophyte tissue never contributed more 99

than 15% to the composition of the faecal material of either species and was usually around 8%.

Cocconeis was the most represented diatom genus in the faeces of both species, and Gomphonema and Epithemia each contributed more than 5% of diatom numbers overall (Appendices 6.1b and 6.2b). Rhoicosphenia. had relatively representation in the faeces of P. tillyardi than P. hendersoni, probably because it was relatively more abundant where the former was collected. On the other hand, Asterionella, a diatom that may be present in the phytoplankton (Stout 1972), was recorded only in the faeces of P. hendersoni. P. hendersoni was collected from an E. canadensis zone underlying open water in the northern and western sampling areas.

A higher percentage of individual P. tillyardi larvae than P. hendersoni larvae had each of the major food categories in their faeces (diatoms:- P. tillyardi 98.7%, P. hendersoni 76.7%; detritus:- 98.7%, 76.7%; macrophyte:- 66.2%, 56.2%; filamentous algae:- 87.8%, 69.9%) •

Differences in food habits between the two species of Paroxyethira, in Lake Grasmere, were probably related to differences in available food in the respective collection areas, rather than selective feeding.

Most Hydroptilidae make exclusive use of filamentous algae by piercing and sucking the cell contents (Nielsen 1948, Wiggins 1978, Mackay & Wiggins 1979). However, Nielsen (1948) observed that the structure of the mouthparts did not exclude the possibility that solid food could be ingested. Of five genera (Agraylea, Hydroptila, Oxyethira, orthotrichia and Ithytrichia) he examined, the first four normally fed on the sap of filamentous algae, but oxyethira was observed (in the laboratory) ingesting delicate filamentous algal threads, and diatoms were also found in its alimentary canal. Ithytrichia, on the other hand, fed exclusively on solid food, viz., very small diatoms. The structure of the mouthparts suggests that they are used to scoop up periphyton from surfaces. Nielsen (1948) did not indicate whether Oxyethira ingested filamentous algal cells or diatoms in nature, but it was interesting that this genus was the only one of the five that he found in acid waters and water where filamentous algae were not present. Ithytrichia was restricted to fast-flowing water also where filamentous algae did not grow. 100

Soszka (1975b) analysed the gut contents of 120 hydroptilids of the genera Agraylea, Hydroptila, Orthotrichia and Oxyethira and found that periphytic algae (diatoms 62%, other periphytic algae 35%) were the main food. In only one case was a small fragment of macrophyte tissue found.

Disney (1972), in a most unusual record, has observed larvae of Orthotrichia in Africa sucking the fluids from the eggs and immobile pupae of black flies (Diptera: Simuliidae).

The food habits of New Zealand Hydroptilidae had not been examined quantitatively previously. Cowley (1978) observed that Oxyethira albiceps was common where filamentous algae abound, sucking the contents from cells. Although I did not examine the gut contents of O. albiceps quantitatively, whole mounts of larvae on microscope slides contained fragments of cell walls of filamentous algae, suggesting that feeding was not solely by piercing and sucking, although the guts never appeared to be full of solid material as was the case wi th the pal:oxyethira species.

Cowley (1978) noted that the mouthparts of paroxyethira hendersoni were very similar to those of O. albiceps implying, perhaps, similar food habits, although this was not stated. The present work, however, confirms the view of Leader (1970) who said "it should be noted that the normal food of hydroptilid larvae is generally thought to be the sap of algae, feeding being accomplished by piercing single algal cells and sucking out the contents, but the guts of paroxyethira spp. are always found to be full of diatoms, and the mouthparts appear to be better adapted for sweeping up a diatomaceous carpet than for biting cell walls".

(2) Hudsonema amabilis (Trichoptera: Leptoceridae). Caddisfly larvae of the family Leptoceridae have been recorded as carnivorous (Hicken 1967, Winterbourn 1971a,Satija 1974, Resh 1976, Mackay & Wiggins 1979), phytophagous (Mackay & Wiggins 1979), detritivorous (Resh 1976) and omnivorous (Slack 1936, Babington 1967).

The overall composition of the faecal material of H. amabilis (instars 2-5) in Lake Grasmere was indicative of omnivorous food habits (Fig. 4.2). Diatoms accounted for a reasonably constant percentage (30 - 40%) of the food of ins tars 2 to 5 (Fig. 4.3) and the major changes with increasing age were the reduction in the detrital component and the increased presence of macrophyte and prey items. Greater detrital feeding in early instars has been documented for many herbivore- 101

detritivore species (Coffman, Cummins & Wuycheck 1971) and a shift from algal or detrital feeding to predation has been demonstrated in the later instars of several caddisflies (Winterbourn 1971b,Wiggins 1977).

10

Fig. 4.3 Food habits of the last four larval instars of Hudsonema amabilis, in terms of major food categories, on a projected area basis (from faecal analyses) (September 1976 - November 1977). Number of larvae examined: 2nd instar = 11, 3rd instar 18, 4th instar = 50, 5th instar = 66. N.B. greater numbers of larvae were examined non-quantitatively (see Tables 4.4 and 4.5) .

. O-L--~~~~~~ mstar 2 3 4 5

Almost all H. amabilis larvae examined had ingested diatoms and detritus and most had fed on filamentous algae (Table 4.5). The number of larvae that had fed on macrophyte tissue and animal food items increased progressively from the second to the fifth instar.

A greater range of prey items was recorded in the faeces of fifth instar larvae than in those of earlier stages (Table 4.4). The most notable change between instars was the increased predation on Potamopyrgus antipodarum and mites. These data suggest that P. antipodarum was the most preyed upon species, and this may be because it was the most common of the potential prey in the habitat (Table 3.9, p. 33). H. amabilis larvae have been observed feeding upon P. antipodarum in the laboratory (see Wilson 1980), on chironomids by Babington (1967), and on small mayflies and P. antipodarum by Cowley (1978).

P. antipodarum of increasingly larger size may be eaten by later instars of H. amabilis (Fig. 4.4). Ninety-six percent of snails eaten by fifth instar larvae were greater than 4 rom tsh (cf. 36% of those eaten by fourth ins tar larvae). Wilson (1980) found in the laboratory that third and fourth ins tar larvae only ate snails of greater than 3 mm tsh and that second instar larvae did not eat P. antipodarum. My data is in general agreement with that of Wilson (1980), although I recorded 102

Table 4.4 Percentage of Hudsonema amabilis larvae in instars 2 5 whose faeces contained various prey items (September 1976 November 1977). N = number of larvae examined.

Instar category 2 3 4 5

N 11 18 87 121 Nematoda 1.6 Oligochaeta 5.6 4.6 5.7 Cladocera Graptoleberis testudinaria 1.6 Bosmina meridionalis 1.6 Copepoda Eucyclops serrulatus 5.6 2.3 Chironomidae 9.1 2.3 Trichoptera Paroxyethira spp. 1.2 2.5 Acarina Piona uncata exigua 1.2 3.3 Hydrozetes lemnae 5.6 5.8 9.0 Trimalaconothrus novus 2.3 3.3 Mollusca Potamopyrgus antipodarum 5.6 23.0 31.2 Unidentified exoskeleton 5.6 1l.5 10.7

P. antipodarum remains in the faeces of only one third instar H. amabilis larva, and the remains of several snails of less than 3 mm tsh in the faeces of third, fourth and fifth instar larvae. Wilson (1980) postulated that fourth and fifth instar H. amabilis larvae could not feed on snails 6f total shell height less than 3 mm and 4 mm respectively because their heads were too wide to enter the snail's shell aperture to get behind the operculum. Unfortunately, this argument is false due to errors in her head width measurements. Final ins tar larvae (mean head width about 0.70 mm) are capable of entering the aperture of P. antipodarum shells as small as 1.0 - 1.9 mm tsh (minimum aperture diameter 0.803 mm (from Wilson 1980)). Therefore, whereas H. amabilis larvae may retain the ability to ingest small snails, they may find large snails a better proposition.

The overall pattern of food habits of each instar was reflected in the seasonal data (Appendix 6.3). The ranking of major food categories in individual months usually was similar to the overall 103

6 n :: 37 0 0 0 o 0 e 0 0 00 o 0 000

E 0 00 00 0 E 4 ooe ..c~ O"l (]) eo 0 ..c

~

~ (]) ..c 0 0 (j) 2 ...... ru 00 0 -I-" 0 00

o 0.5 1.0 1.5 hind tibia length (mm)

Fig. 4.4 Sizes (calculated from radula teeth) of Potamopyrgus antipodarum whose remains were found in the faeces of Hudsonema amabilis larvae of various sizes (September 1976- November 1977). Solid circles ~ two records.

ranking for each instar. Few clear seasonal trends were evident, although from November to January the representation of prey items in the faeces of fifth instar larvae was greater than at other times, suggestive of increased emphasis on higher quality food prior to pupation (see Anderson & Cummins 1979).

Cocconeis was the most common diatom (63 - 81%) in the faeces of the last four instars of H. amabilis. No other diatom genus contributed more than 10% by numbers to the overall diatom category for any instar, although some genera were more prominent in certain months (Appendix 6.4). Rhoicosphenia and Epithemia (instars 2 - 5), Asterionella(instars 2 and 5), Cyclotella (instars 2, 3 and 5) and Gomphonema (instars 2 and 3) were all well represented at various times of year (Appendix 6.4) . 104

Table 4.5 Percentage of Hudsonema amabilis larvae in ins tars 2 - 5 whose faeces contained each of the major food categories

(September 1976 - November 1977) (N = number of larvae examined) .

Major food category Instar 2 3 4 5 Overall

N 11 18 87 121 237 Diatoms 90.9 94.4 98.9 100.0 98.7 Detritus 100.0 94.4 98.9 97.5 97.9 Macrophyte 27.3 66.7 73.6 91.7 80.2 Filamentous algae 72.7 77.8 87.4 84.3 84.4 Animal 9.1 27.8 48.3 62.8 52.3 S.R.T's 27.3 44.4 18.4 22.3 22.8

A study of the larval food habits of H. amabilis in Lake Rotorua (North Island) by Babington (1967) showed that the larvae had catholic feeding habits, a feature that is confirmed by the present work. The gut contents of ten larvae she collected in September were dominated by animal remains (including entire tiny caddis larvae, insect head capsules, thoracic plates, wings and legs), with egg cases, macrophyte tissue and algae also present. In the guts of a further ten larvae collected in March, however, diatoms and filamentous algae 'were most prominent, followed by macrophyte tissue and egg cases. No animal fragments were found. Babington (1967) attributed this seasonal change in food habits to a change in available food in the habitat. The same seasonal pattern of change in food habits was not observed in the present study (Appendix 6.3). Animal remains were present in the faeces of larvae collected in most months (the exceptions being second and third ins tar larvae when low numbers of larvae were collected), but could not be referred to as the dominant food category at any time. However, the conclusion that availability is the key to trophic relationships is supported, for example, by the species composition of the prey taken, and the wide range of diatom genera eaten.

(3) Triplectides cephalotes (Trichoptera: Leptoceridae). The faecal material of second to fifth instar T. cephalotes was dominated by macrophyte tissue (Fig. 4.2 and Fig. 4.5) although diatoms, detritus, filamentous algae, prey and S.R.T's were also recorded. The food 105 habits of second and third instar larvae were very similar (Fig. 4.5) but the faeces of only two second ins tar larvae were examined. Macrophyte tissue was ingested to a relatively greater degree by fourth and fifth instar larvae, mainly at the expense of detritus and filamentous algae.

10

Clo U c: III Fig. 4.5 -0 c: :::J Food habits of the last four instars of .0 5 III Triplectides cephalotes, in terms of Clo major food categories, on a projected :> ..- area basis (from faecal analyses) .!9 ....Clo (September 1976 - November 1977). Number of larvae examined: 2nd ins tar = ~ 0 2, 3rd instar = 8, 4th instar 23, 5th instar == 31.

O-.~-+~~~~~~ Instar 2 3 4 5

Table 4.6 Percentage of Triplectides cephalotes larvae in instars 2 - 5 whose faeces contained various prey items (September 1976- November 1977). (N = number of larvae examined.)

Instar Prey category 2 3 4 5

N 2 8 23 31 Oligochaeta 4.3 3.2 Cladocera Graptoleberis testudinaria 50.0 8.7 21.6 Bosmina meridionalis 37.5 Trichoptera Paroxyethira hendersoni 3.2 Acarina Hydrozetes lemnae 6.5 Mollusca potamopyrgus antipodarum 6.5 Unidentified exoskeleton 4.3 106

A very limited range of animal food items was found in the faeces of T. cephalotes larvae (Table 4.6) and cladocerans were the only group represented in the faeces of all instars examined. The data for the second and third instars should be regarded with caution due to the small sample size. Bearing this in mind, it is likely however that the carnivorous habit assumes a greater role in progressively later instars. As with H. amabilis, the fifth instar had the greatest prey diversity, but with only 29% of larvae examined having prey remains in their faeces (Table 4.7), carnivory may be an irregular occurrence. Eighty-six percent of all prey items found in faeces of T. cephalotes were from fourth and fifth ins tar larvae collected between November and January, but only 59% of the animals examined were collected during the same period (Appendix 6.5). This suggests that animal food was taken to a greater extent in summer and may be another example of increased carnivory prior to moulting into the final instar or pupation (see Anderson & Cummins 1979).

Almost all individuals in instars 2 5 had ingested diatoms, detritus and macrophyte tissue (Table 4.7), the first two a likely consequence of a preference for the last. Filamentous algae were ingested by a progressively smaller proportion of later than earlier instars.

Table 4.7 Percentage of Triplectides cephalotes larvae in instars 2 - 5 whose faeces contained each of the major food categories (September 1976 November 1977). (N number of larvae examined. )

Instar Major food category 2 3 4 5 Overall

N

Diatoms 100.0 100.0 100.0 100.0 100.0 Detritus 100.0 100.0 100.0 100.0 100.0 Macrophyte 100.0 100.0 100.0 96.8 98.4 Filamentous algae 100.0 75.0 69.6 48.4 60.9 Animal 50.0 37.5 17.3 29.0 26.6 S.R.T's 100.0 87.5 52.2 51.6 57.8 107

Interpretation of patterns of seasonal change in the contributions of major food categories (Appendix 6.5) and diatom genera (Appendix 6.6) to the food of T. cephalotes was made difficult by the low numbers of larvae available for examination in some months. In ten out of 18 months, the faeces of three or fewer T. cephalotes larvae were analysed. Macrophyte tissue was the top-ranked contributor to the food of larvae in most months for ins tars 2 to 5 (Appendix 6.5), with detritus (3rd instar), and diatoms (2nd instar) dominant when macrophyte was not. Cocconeis was the best represented diatom in the faeces of most larvae throughout the year (Appendix 6.6). Gomphonema, Epithemia and Rhoicosphenia were next in abundance for both fourth and fifth instar larvae (although not in the same ranking order) .

T. cephalotes has an almost obligatory relationship with aquatic macrophytes. Babington (1967) observed larvae chewing leaf margins of E. canadensis, Potamogeton cheesemanii and P. ochreatus in Lake Rotorua (North Island) . Older leaves, especially of E. canadensis, were preferred and new growth was not eaten. T. cephalotes larvae normally construct their cases from potential food items, and in the macrophyte zones of Lake Grasmere, E. canadensis leaves and stems, and hollow stems of I. alpinus, are most commonly utilised. Sometimes stones or P. antipodarum shells may be incorporated (as noted also by Cowley (1978». At times of food shortage (for example, isolated larvae in the laboratory) larvae may eat their own cases, or those of others.

Babington (1967) found that macrophyte tissue was the dominant food item in gut contents in September (up to 50% by projected area) and March (when filamentous algae were almost equally important). In March, the guts of some individuals were almost completely full of diatoms. _These data are in agreement with the results of this study.

Few trichopteran larvae use green plants for food and case construction (Mackay & Wiggins 1979). The use of plant materials and silk (as opposed to sand and stones) by larvae such as Triaenodes (Moretti 1942, in Mackay & Wiggins 1979) and T. cephalotes (Babington 1967, Cowley 1978, present study), has enabled these larvae to leave the bottom substrate to which they would otherwise have been restricted. None of these larvae can swim without their cases, and Mackay and Wiggins (1979) suggest that the slender plant and silk-based cases serve a hydrostatic or hydrodynamic function in making macrophyte energy resources available.

Phytophagous leptocerids include one of the few larval trichopteran pests; the larva of Triaenodes, which can be a pest of rice crops (Moretti 108

1942, in Mackay & Wiggins 1979). (The only other larval Trichoptera that assume pest status in crops belong to the family Limnephilidae, which is not represented in New Zealand.) Although the opportunity for trichopteran larvae to become pests in New Zealand is limi.ted (because there is little, if any, commercial production of aquatic plants), the use of phytophagous larvae (such as T. cephalotes) in the biological control of nuisance aquatic macrophytes (by artificially increasing larval densities) is a possibility.' B.T. Coffey (pers. comm. to Cowley 1978) has observed large larvae of T. cephalotes on defoliated stems of Lagarosiphon indicating that larvae can reduce the standing crop of the plant.

(4) Nymphula nitens (Lepidoptera: Pyralidae).· The food habits of N. nitens in New Zealand have not been studied previously but it has been assumed, by analogy with the habits of overseas species, that living macrophyte tissue is the main food. Winterbourn and Lewis (1975) suggested that N. nitens and T. cephalotes were "perhaps the only inhabitants of the macrophyte beds that eat the living plants".

In Lake Grasmere, N. nitens larvae are phytophagous (Fig. 4.2). Nearly 60% (by projected area) of faecal material was macrophyte tissue. Detritus (20.0%), diatoms (13.6%), and filamentous algae (6.6%), which contribute to the periphytic coating on plant surfaces, would almost certainly have been ingested while chewing on macrophyte tissues.

Because limited numbers of larvae were available and the size ranges of instars were uncertain, three arbitrarily chosen size classes of N. nitens were considered for food comparisons (Table 4.1).

Macrophyte tissue contributed an increasing proportion to the diet of larvae as they got larger (Fig. 4.6), and comprised 38.0%, 60.5%, and 71.0% of the faeces of larvae of the first, second, and third size classes respectively. The parallel reduction in the contributions of detritus and diatoms resulted from the ingestion by larger larvae of larger fragments of macrophyte (with less periphyton relative to plant tissue than smaller fragments). In addition, larger larvae were able to eat their way into plant stems where they could ingest tissue that was not coated with diatoms 0r detritus.

Seasonal changes in food habits (Appendix 6.7) were difficult to interpret due to the difficulty of separating seasonal effects from individual variation between larvae since, in most months, sample sizes were small (see Chapter V, p. 132). Filamentous algae were most prominent in August/September and diatoms least represented from August 109

to November in the faeces of all three size classes (Appendix 6.7a). Detritus showed marked seasonal variation in contribution but no consistent pattern. Macrophyte tissue made a consistently large contribution to the ingested food of each size class in most months. Lowest values were found in June (size groups 2 and 3), July and September (size group 1).

Cocconeis was the dominant diatom in the faeces of N. nitens throughout the year. Diatom species richness was lowest, and the dominance of Cocconeis most marked, in the faeces of larvae in the largest size class (Appendix 6.7b).

100

Fig. 4.6

c:: Food habits of three size classes of 0 :~ Nymphula nitens, in terms of major food (/) 0 50 categories, on a projected area basis 0- E (from faecal analyses) (July 1976 - 0 u August 1977). See Table 4.1 for size ranges of the size classes. Number of ~ 0 larvae examined: size class 1 8, size . " .. ,,' ,,'. , .. , .... . ' .. ... , ' .. " ..... class 2 = 28, size class 3 = 7 . .... 0 0 00 00 00 o 000 size group 1 2 3

Table 4.8 Percentage of Nymphula nitens larvae in three size classes whose faeces contained each of the major food categories (July 1976 - August 1977) (N number of larvae examined).

Major food category Size class 1 2 3 Overall

N 8 28 7 43 Diatoms 100.0 100.0 100.0 100.0 Detritus 100.0 100.0 100.0 100.0 Macrophyte 100.0 100.0 100.0 100.0 Filamentous algae 87.5 85.7 85.7 86.0

Cocconeis was the most common diatom on macrophytes and one of the first species to colonise new surfaces (see p. 98). The generic llO

composition of diatoms in the faeces of larvae of the third size class suggests that these larvae were feeding on shoot tips of macrophytes. This was confirmed by observation; small larvae were distributed more evenly over the plants but large larvae were invariably found where there was new growth.

All larvae examined had ingested diatoms, detritus, and macrophyte tissue (Table 4.8). A similar proportion (about 86%) of larvae of each size class had fed upon filamentous algae.

Cummins (1973) noted that there was a paucity of good quantitative data on the food habits of aquatic Lepidoptera, despite their being widely distributed in freshwaters. Welch (1916) observed that larvae of Nymphula maculalis fed on yellow water-lily (Nymphaea americana) in Douglas Lake (Northern Michigan, U.S.A.) and Berg (1941) found larvae of Acentropus niveus only at the tips of shoots (or near the tips) of Elodea canadensis and Ceratophyllum sp. where they fed on "the youngest and tenderest leaves". Also, larvae construct "houses" with silk threads, from pieces of leaves of the food plant. These observations are consistent with those of the present study, where larvae were found most commonly in the shoot tips of Myriophyllum propinquum and, when densities were high (e.g., March 1978), on E. canadensis. At high larval densities, N. nitens can have a marked defoliating effect, and may be the most promising indigenous species for use in biological control of nuisance aquatic macrophytes.

Soszka (1975b) provided the first quantitative data on the diet of aquatic Lepidoptera (Paraponyx stratiotata and A. niveus). Both species had identical food habits and, as for N. ni tens , all larvae examined (n = 100) primarily on the tissue of vascular plants. The percentage composition of gut contents (on a projected area basis) was approximately 53.0% macrophyte, 17.6% diatoms, 17.6% other algae, 5.9% detritus and 5.9% invertebrates (viz., fragments of Cladocera, Oligochaeta, and Rotatoria) (determined from Fig. 2, p. 398, Soszka 1975b). Although no animal fragments were recorded in the faeces of N. nitens, the remaining food categories were similar to those recorded by Soszka (1975b). Differences in percentage composition were probably due to differences in available food.

In experiments with artificial substrates, Soszka (1975b) found that lepidopteran larvae were the only macroinvertebrates in Mikolajskie Lake that did not occur on plastic plants. This emphasises the importance of fresh macrophyte tissue as an essential food of aquatic Lepidoptera. III

(5) xanthocnemis zealandica (Odonata: Coenagrionidae). Odonatans provide perhaps the best examples of how similarity in morphology confers similarity in food habits. The presence of the modified labium in all larval Odonata is considered sufficient evidence to conclude that all species are predaceous, even though the food habits of only a few species have actually been studied (Cummins 1973). Odonatan larvae are among the more convenient predators to investigate for field-consumed prey because the faecal material is passed in pellets enclosed by a peritrophic membrane.

Methods of faecal pellet analysis have been used to study the prey of zygopteran damselfly) larvae (Lawton 1970a, Pearlstone 1973, Thompson 1978a & b) and anisopteran dragonfly) larvae (Corbet 1957, Pritchard 1964, Staddon & Griffiths 1967). Other studies have utilised gut content analyses (Chutter 1961, Macan 1964, Fischer 1966, 1967, Crumpton 1979).

A wide range of prey types are taken by odonate predators, but certain groups have not been recorded in gut contents or faecal pellets. These include adult Coleoptera and larval Trichoptera (Chutter 1961, Lawton 1970, Pearlstone 1973, Thompson 1978a & b), and corixids have been noted as prey of zygopteran larvae only by Crumpton (1979). The presence of remains of hydroptilid caddisfly larvae in faecal pellets of X. zealandica (present study) is thus the first record of predation on Trichoptera. Serological studies have shown that damselfly larvae may prey upon triclads (Davies 1969), but these (and coelenterates and leeches) are likely to be the only prey species completely overlooked by faecal pellet (or gut content) analysis (Davies & Reynoldson 1971). It was not possible to eliminate this minor source of error given the time and resources available, and none of these potential prey were common in the microhabitat of X. zealandica in Lake Grasmere.

Few studies of the trophic relationships of Odonata have recognised non-animal 'food' materials in gut contents or faecal pellets. However, Thompson (1978a) recorded plant detritus (which appeared as a fine brown material) in faecal pellets, some of which he contended was derived from the gut contents of prey organisms and the remainder picked up accidentally during prey capture.

As expected, the faeces of Xanthocnemis zealandica were dominated by animal remains, but all of the other major food categories (Fig. 4.2) were also recorded. It is probable that non~animal food items were either ingested accidentally when striking at prey, or were derived from 112 prey gut contents, Diatoms, detritus, and prey items were found in faecal pellets of all X. zealandica larvae examined (Table 4,9),

Table 4.9 Percentage of Xanthocnemis zealandica larvae in three size classes whose faecal pellets contained each of the major

food categories (September 1976 - November 1977) (N = number of larvae examined).

Size class Major food category 1 2 3 Overall

N 26 53 47 126 Diatoms 100.0 100.0 100.0 100.0 Detritus 100.0 100.0 100.0 100,0 Macrophyte 73.1 88.7 83.0 83.3 Filamentous algae 84.6 81.1 72.3 78.6 Animal 100.0 100.0 100.0 100.0 S.R.T's 1l.5 39.6 29.8 30,2

Since the instars of X. zealandica could not be determined, and larval densities were low (which limited sample sizes), only three size classes were chosen for the purpose of faecal analyses (Table 4.1).

The third (= largest) size class approximated the final ins tar (see Deacon 1979).

There was very little difference in food habits (in terms of major food categories) between the three size classes of X. zealandica larvae examined (Fig. 4.7). Differences in food habits between size classes were evident when the composition of the prey category was considered (Table 4.10). A high proportion of larvae of all sizes fed on Cladocera (mainly Graptoleberis testudinaria) and Oligochaeta. Mites (especially Hydrozetes lemnae and Piona uncata exigua) , hydroptilid caddisfly larvae and, especially potamopyrgus antipodarum had greater representation in faecal pellets of larger X. zealandica larvae. Conversely, chironomid larvae were found more often in the pellets of smaller larvae.

Final instar X. zealandica larvae apparently spend most of their time in a head down posture close to the bottom of stems (the "hunting" position) (Rowe 1978). Nevertheless, some larvae may be found at the tops of stems (facing up) and Rowe (1978) suggested that this was either in preparation for emergence or a method of avoiding intraspecific conflict. 113

Table 4.10 Percentage of Xanthocnemis zealandicd larvae in three size classes whose faecal pellets contained various prey items (September 1976 - November 1977) (N= number of larvae examined).

Size class category 1 2 3

N 26 53 47 Oligochaet':l 65.4 73.6 76.6 CladQcel:a Graptoleberis testudinaria 73.1 75.5 61. 7 Bosmina mer id.iona1.is 26.9 17.0 10.6 Simocephalus vetulus 4.3 Copepoda Eucyc.lops se.rru.l.a tus 2.1 Ostracoda 1.9 Chironomidae Tanypodinae (Gressittius antarcticus) 3.9 Orthocladiinae 15.4 9.4 6.4 'ranytarsini (TanytarSl.lS vespertinus) 1.9 Trichoptera Hydroptilidae 1l.5 22.6 21. 3 Unidentified 3.9 Acarina Hydro2etes Jemnae 11.5 22.6 27.7 Tr.imalaconothrus novus 1.9 2.1 Fiona uncata exigua 1l.5 20.8 34.0 Mollusca E'otamopyr:gtlS antipodarum 1.5.1 42.6 unidentified exoskeleton 11.5 7.6 6.4

100

a. u C tU U Fig. 4.7 c .E 50 Food habits of three size classes of cO Xa.nthocnemis zealandica, ill terms of major food categories, on a projected area basis (from faecal pellet analyses) (September 1976 - November 1977). Number of larvae examined: size class 1 = 26, size class 2 : 53, size class .3 = 47. O-~--··~~+--4-~··r 51Zc grou 114

The food habits clf final ins tar la.rvae recorded here (viz., the high contributions of oligochaetes, benthic cladocerans, Acarina, and P. antipodarum (see Chapter III. p. for discussion of the habits of these groups) are consistent with the abundance of potential prey in the collection area (Table 3.9, N E 2, W E 2, p. 33).

crumpton (1979) studied the food habits of X. zealandic:a in two ponds near Chris1:church (South Island, New Zealand) but did not separate larvae of different sizes. Oligochaeta, Ostracoda, Cladocera and chironomid larvaE~ were most frequently represented in crop contents. Predation on tiny zygopteran larvae (i.e., cannibalism) was recorded in 9.4% of crops examined. I found no instances of cannibalism in an examination of 126 faecal pellets, and Rowe (1978 and pers. comm.) 1 in many hours of observa.tion, has observed cannibalism only in crowded transport containers. Work on related species overseas shows that the incidence of cannibalism is very low. Lawton (1970a) identified

2, 000 prey items from pyr.r:hosoma nymph111a and found only two zygopt(~ran remains, and Chutter (1961) found only one instance (in 600 crops) where a zygopteran (a different species) had been eaten by pseudagrion.

X. zealandica larvae ingested P. antipodarum of all sizes available in Lake Grasmere (range 1.48 - 5.24 tn.'1l tsh determined using the tsh versus length of inner marginal tooth regression, p. 93).

Final instar larvae (= size class 3) showed a preference for larger snails than did X, zealanciica of size group 2, bu't both siz(~ classes retained the ability to feed on small snails (Fig. 4.8). X. zealandica la.rvae smaller than 2 rom h. w. v.Jere not recorded as predators of P. antipodarum. Crumpton (1979) did not record P. ant.ipoda.rum in crop contents of X. z;ealandica despite its presence (few to common, all year round) in the habitat (Crumpton 1978) .

Seasona.l changes in the composition of faecal pellets by major food categories I and the geneo.c composition of the diatom cateqory are given in Appendices 6.8 and 6.9. Since non-prey food items may be regarded as somewhat incidental to the nutrition of damselfly larvae, seasonal changes in the representations of prey items only, are considered.

The overall pattern of prey taken (Table 4.10) for each size class was reflected in seasonal data (Table 4.11) with Oligochaeta, Cladocera, Acarina and P. antipodarum most represented in faecal pellets. No marked seasonal changes were evident, reflecting year-round availability of the prey items recorded (see Appendices 2.1 and 2.10). 115

6 n:: 24

0 00 0

E 8 E 0 0 4 0 0 4-' ..c. 0 0 0 en 00 OJ ..c. 0 OJ ..c. 0 (/) 0 0 ttl 2 4-' 0 +-" 0 0 0 0

0

size 0 class 1 2 3 0 2 4 head width (mm)

Fig. 4.8 Sizes of Potamopyrgus antipodarum'(calculated from radula teeth) whose remains were found in the faecal pellets of Xanthocnemis zealandicalarvae (September 1976 - November 1977) •

(6) Potamopyrgus antipodarum (Gastropoda: Hydrobiidae). P. antipodarum is the dominant mollusc, if not the dominant macroinvertebrate, in many New Zealand lakes (Forsyth 1975a,Stout 1975b, Winterbourn & Lewis 1975). Nash (1974) made a preliminary study of feeding and assimilation of P. antipodarum using radio tracer techniques and Winterbourn (1970a) observed that epipelic and epiphytic periphyton was the main food. Marshall (1974) stated that P. antipodarum was an epipelic detritivore and these observations were confirmed by Towns (1976).

These statements on the food habits of P. antipodarum were confirmed for Lake Grasmere animals by faecal and observations made during this study. P. antipodarum, in aquaria, were seen browsing over plant surfaces and on periphyton growing on the glass.sides. Feeding trails were visible on the sides of (where the periphytic coating had been rasped off) and there was no visual evidence of selective feeding. Table 4.11 Seasonal occurrence of prey items in faecal pellets of Xanthocnemis zealandica (September 1976 - November 1977) 1 = present in faecal pellets of size class li 2 = present in faecal pellets of size class 2; 3 = present in faecal pellets of size class 3; - = absent; * larval size classes not present. (See Table 4.1 for size ranges of size classes and Appendix 6.8 for numbers of larvae examined,)

Sampling dates Prey category 1976 1977 2 Sep •. 2 Nov. 2 Dec. 20 Jan. 8 Apr. 20 Jun. 21 Nov.

Oligochaeta -,2,3 1,2,3 1,2,3 1,2,3 *,2,3 1,-,3 Cladocera Graptoleberis testudinaria 1,2,- -,2,3 1,2,3 1,2,3 *,2,­ 1,2,3 *,*,3 Bosmina meridionalis 1,2,3 1,2,3 -,-,3 1,-,­ *,-,3 Simocephalus vetulus -,-,- , , I , -,-,3 *,-,- -,-,- *,*,3 Copepoda

Eucyclops serrulatus , , , I I , -,-,3 * I -I - , , * I * , - ostracoda , , , , , I -1-'- *,2,- I I * , * I - Chironomidae Tanypodinae 1,-,- , I -,-,- , , * , -, - , , * I * , - Orthocladiinae 1,2,- -,2,- -,-,3 * , - , - -,2,3 * , * I - Tanytarsini , I , , , , * I -I - -,-,- * , * I - Trichoptera

Hydroptilidae , , 1,2,3 1,2,3 -,-,3 * I -I - -,2,3 *,*,3 unidentified 1,-,- -,-,- -,-,- -,-,- * , -, - , , * I * , - Acarina Hydrozetes lemnae 1,2,3 1,2,3 -,2,.3 *,2,- -,-,- *,*,3

Trimalaconothrus novus , I I , , I * , -I - -,-,3 * I * I - piona uncata exigua 1,:-,3 1,2,3 -,-,3 *,2,- -,2,3 *,*,3 Mollusca Potamopyrgus antipodarum -,2,3 -,2,3 -,2,3 -,2,3 *,2,- -,-,3 *,*,3 unidentified exoskeleton 1,2,3 1,-,3 I I -,2,- * , - , - 1,-,3 * , * , - 117

The faecal contents of P. antipodarum (in terms of major food categories) were similar to those of the two Paroxyethira species (Fig. 4.2). Diatoms were the main food items, followed by macrophyte fragments, detritus, filamentous algae and S.R.T's. On this level, the main differences between these species were the increased representation of macrophyte tissue and the decreased representation of filamentous algae in the faeces of snails compared with hydroptilids. The radula of P. antipodarum is capable of rasping off the epidermal layer of macrophyte tissue, whereas the scooping mouthparts of Paroxyethira spp. (cf. Ithytrichia, Nielsen 1948) seem able only to scrape off the periphytic coating. Also, Nielsen (1948) observed that the structure of the mouthparts of algal piercing and sucking hydroptilids differed little from those that had taken up a browsing habit. These functional differences between the mouthparts of P. antipodarum and the Paroxyethira spp. may account for the observed differences in their faecal composition.

All P. antipodarum whose faeces were analysed had fed upon diatoms and detritus, and most on macrophyte (95.8%) and filamentous algae (81. 3%) • S.R.T's were recorded in the faeces of 35.4% of snails examined (mainly in June 1977 when S.R.T's were most abundant (Appendix 6.10». No animal remains were found in faeces of P. antipodarum. There were marked seasonal changes in the relative contributions of major food items eaten by P. antipodarum (Appendix 6.10d) but, as with most of the other species studied, it is difficult to explain them. Diatoms were the most common food items in all months in which analyses were done, except April 1977 (when macrophyte, detritus and filamentous algae were more common). Detritus and macrophyte tissue were usually next to diatoms in ranking and contributed similar amounts in most months. Filamentous algae comprised less than 12% of the faeces (by projected area) in all months except April 1977 (22.4%).

The generic composition of the diatom category in the faeces of P. antipodarum (Appendix 6.10b) was distinctly different from that of any of the other species studied. Rhoicosphenia and Cocconeis made approximately equal overall contributions and Epithemia had greater representation in the faeces of P. antipodarum than in those of the other animals. Since most P. antipodarum were collected from short- stemmed I. alpinus beds, in the eastern sampling zone of Lake Grasmere, it is possible that they may have been feeding closer to, or on, the bottom (on periphyton-covered rocks or beech tree debris). Rhoicosphenia and Epithemia are two of the most common diatom genera in Middle Bush Stream, Cass (Davis & Winterbourn 1977) where such substrates may be 118

found I so it is possible that these diatom genera also were more abundant in Lake Grasmere in the habitat where these P. antipodarum were feeding.

(7) contents of Chironomidae. During the course of taxonomic studies on New Zealand chironomids, many larvae were mounted on slides usually without treatment with KOH so that their gut contents were often visible. Since little is known about the food habits of the New Zealand species, I considered it useful to document these observations.

Tanypodinae Larvae of the subfamily Tanypodinae generally are considered to be predaceous (Cummins 1973), although Davies and McCauley (1970) recorded the presence of algae as well as animal remains in digestive tracts of Tanypodinae from Marion Lake (Canada).

Large numbers of Gressittius/Macropelopia were examined from lakes throughout the South Island of New Zealand, and in most cases, diatoms and detritus dominated the gut contents. In many larvae, very

large diatoms (e.g., stauroneis (?) up to 2.8 mm long) were present. Other chironomids (e.g., Chironomus zealandicus) and Cladocera (especially Graptoleberis testudinaria and Bosmina meridionalis) were the main animal prey items seen.

The gut contents of Ablabesmyia mala invariably contained prey items including chironomids (notably Tanytarsus vespertinus in Lake Grasmere), Cladocera (especially G. testudinaria in Lake Grasmere), and Copepoda.

Orthocladiinae Orthocladine chironomids are normally detrital-algal feeders (Cummins 1973). Observations on the gut contents of Syncricotopus pluriserialis, Cricotopus spp., Rheocricotopus Spa Orthocladiinae B) and Orthocladiinae C (see Chapter VI) supported this contention.

Chironominae Algal-detrital feeders, leaf-miners and wood-dwellers are known in this subfamily (Cummins 1973).

Tribe Chironomini Harrisius pallidus, a wood-dwelling species, is closely related to the genus Stenochironomus, whose members are leaf-miners or wood- 119

dwellers (A. Borkent pers. COmIDo) 0 Digestive tracts of Ho pallidus

were always full of wood fragments 0 Paucispinigera spp. may also feed on wood, but in fine particulate form, from amongst sediments or on the surfaces of logs. Guts of this genus may also contain large quantities of inorganic material, as well as detritus and diatoms.

The of Chironomus zealandicus examined were normally full of fine particulate organic and inorganic material.

Tribe Tanytarsini The gut contents of Calopsectra funebris, Tanytarsus vespertinus and Corynocera sp. were dominated by detritus and diatoms, with large quantities of mineral particles in some larvae (e~pecially C. funebris).

121

CHAPTER V

LIF HISTORY INFORMATION ON SELECTED INSECTS

5.1 INTRODUCTION

Although the study of life-histories of aquatic insects in New Zealand has advanced significantly over the last ten years or so (e.g., Winterbourn 1966, 1974, 1978, Norrie 1969, Michaelis 1973, Crosby 1975, Hopkins 1976, Winterbourn & Davis 1976, Towns 1976, Cowie 1980) most work has concerned stream insects. Life-history studies by Babington (1967) (three leptocerid caddis flies), Deacon (1979) and Crumpton (1979) (Odonata), and Young (1970) (Hemiptera), are among the few investigations of pond- or lake-dwelling insects. Brief life-history notes are provided, usually as part of taxonomic studies, for a number of other species by Forsyth (1971, 1979), Forsyth & McCallum (1978a & b) (Chironomidae), Leader (1972) (Hydroptilidae), and Cowley (1978) (Trichoptera) .

Life-history information interrelates with, and is a necessary complement to, the study of the taxonomy of aquatic insects (Oliver 1979) . For example, knowledge that certain immature stages and adults belong to the same species aids in the search for distinguishing taxonomic characters in one or all of the life stages. In addition, a life- history study may reveal morphological or behavioural differences that result in the recognition of two species where only one was thought to be present (Oliver 1979). Information on the seasonality of abundance of various life stages is important in the elucidation of food webs and trophic int:errelationships (Cummins 1973, Chapter IV of present study) since different stages may have different food habits. There is evidence also that interspecific competition may be reduced by life­ history asynchrony such that maximum numbers of major feeding stages (i.e., final instar larvae) of potential competitors are separated temporally to some degree (see Winterbourn 1971a).

This study is concerned primarily with the life-histories of the insects whose feeding relationships were also studied (see Chapter IV), although information was collected on a number of other species as well. Since I was most concerned with aquatic insects in their association with macrophytes, the emphasis was on the larval stages. Information obtained on other life-history stages (especially the egg) is more limited. Nevertheless, for most species it was possible to determine the main life­ history patterns. 122

5.2 METHODS

Aquatic stages of insects were collected in quantitative samples and non-quantitative hand-net collections (see Chapter III for details of sample collection and processing) . Limited numbers of adult insects were obtained by sweep-netting lakeside vegetation (Appendices 5.1, 5.3, 5.5 and 5.7) particularly at the southern end of the lake, and by light-trapping (Appendices 5.2, 5.4, 5.6 and 5.7).

Maximum head width or hind tibia length were measured (on material preserved in 70% alcohol) in order to assign individual larvae to instars or size classes. Maximum head widths (hw) were measured for pycnocentrodes aureola, Oecetis unicolor and Nymphula nitens at x40 magnification, and at x25 for Xanthocnemis zealandica, using a binocular microscope fitted with a linear eyepiece micrometer. It was inconvenient to measure head widths of the two larger leptocerid caddis larvae (Hudsonema amabilis and Triplectides cephalotes) since their greatly elongated meta-thoracic limbs hampered correct orientation. As preserved larvae tend to lie naturally on their sides, with their elongated limbs oriented approximately in the focal plane of the microscope, the length of the hind tibia (the longest leg segment) was measured instead. As a check on reliability for instar determination, and to enable comparisons to be made with studies in which head width was used, both structures were measured for a number of individuals, including some collected during different seasons.

Linear regressions of hind tibia length (htl) versus head width (hw) were fitted by the least squares method for 69 H. amabilis (Fig. 5.1) and 39 T. cephalotes (Fig. 5.2) larvae. Measurements of hind tibia length provided a convenient and unambiguous method for assigning individuals of these species to their appropriate instars, and being between 6% and 55% greater than head width (for smallest and largest larvae respectively) for H. amabilis, and 22% and 69% greater for T. cephalotes, there was less error involved in their measurement with an eyepiece micrometer. The relationship betweenhtl and hw for H. amabilis has been substantiated by a more intensive study of a lowland population in the Avon River, near the Student Union building, on the Ilam campus of the University of Canterbury (C.L. McLay and K.W. Fraser pers. comm.).

Since the number of instars of X. zealandica could not be determined precisely (but is likely to vary between 12 and 14, R.J. Rowe pers. comm.), its life-history pattern was demonstrated by 123

1. h.t.t.=1.61 h.w. - 0.052 mm

r = 0.98 o Fig. 5.1 Hind tibia length (htl) versus head width (hw) for Hudsonema amabilis with .!!! ..a least square regression ·z I line and equation. "0 ~0.5

o III

.=> 1 record N=69

0.4 0.6 0.8 head width (mm)

2.0

h.Ll.: 1.763h.w 0.101 mm

r = 0.99 Fig. 5.2

~ E Hind tibia length (htl) .5 versus head width (hw) for .s:::; Triplectides cephalotes 0. c with least square ~ regression line and .0 equation.

I "0 c.: .s:::;

(I} >1 record

o 1.0 head width (mm) 124

separating monthly collections of larvae into component cohorts (see Deacon 1979). Cohorts (or year classes) of larvae comprised individuals of similar size range, which could sometimes be distinguished by inspection of monthly size-frequency data. At other times, it was necessary to use Taylor's method for polyrnodal analysis (Taylor 1965) to separate cohorts (see Deacon 1979).

5.3 RESULTS AND DISCUSSION

5.3.1 Hudsonema amabilis (Trichoptera: Leptoceridae)

Measurements of larval hind tibia lengths formed five groups indicating the presenc'e of five larval instars (Fig. 5.3). Instars were distinctly separate in most months, although the combination of htl measurements from all larvae collected between April 1976 and March 1978 slightly obscured the separation between instars, suggesting that the size limits of the instars may be somewhat flexible. Size variation within each ins tar increased with age.

200 3 4 N:o 2642

2

5 n 100

O+----L.----.-----r-----.----.-----r=~_,r_--_.----_.--~== o 0.5 1.0 hind tibia length (mm)

Fig. 5.3 Size frequency distribution of hind tibia lengths of Hudsonema amabilis larvae (all data combined, April 1976 - March 1978).

The numbers of larvae taken on each sarnp~ing date between July 1976 and March 1978, percentage of larvae in each instar, and mean size of larvae in the most abundant instar are shown in Fig. 5.4. First instar larvae were taken from January to April, although one was found in September (1976). Growth was rapid between March and June so that most larvae overwintered in the fourth instar. As lake temperatures rose the following spring, larvae entered a period of rapid growth culminating in emergence. 125

N 70 70 142 54 84 25 248 71 23 423 577 435 155 121 43 215 A

p p p p p p p p

" II

" .. ~ OJ) c :e.!l! 3 , '0- 2 .5 .J:: 1 °JJASON D J J A 5 0 N D 1976 1977

Fig. 5.4 Life-history pattern of Hudsonema amabilis (July 1976 March 1978) showing the percentage of larvae in each instar, and periods of adult (A) and pupal (P) abundance. Solid circles indicate the mean sizes of larvae in the dominant instars. N = number of larvae collected.

First and second ins tar larvae were found mainly on the stony bottom in macrophyte-free, shallow (less than 0.3 m deep) water near the lake edge whereas many later ins tar larvae also occurred on macrophytes. Most pupae were seen attached to rocks and submerged beech tree logs, although some were attached to macrophytes. Since my samples were collected primarily from macrophytes, early instar larvae and pupae are probably under-represented.

Adults of H. amabilis were the most common leptocerid caddisflies

taken in hand-net and light~trap collections and were recorded in November, December, and January (Appendices 5.3 and 5.4). The presence of pupae and final instar larvae (Fig. 5.4) indicate a flight period extending from early November to early March at Lake Grasmere. McFarlane (1977) recorded adults in light-trap collections from the Winchmore Irrigation Research station, about 140 m a.s.l., in inland mid-Canterbury, from 2 November to 26 April whereas around Auckland, in the North Island, adults have been collected from early August to late April (Cowley 1978).

The life-history pattern observed at Lake Grasmere, and the size ranges of the instars (Fig. 5.1) are essentially the same as those found in McLay and Fraser's larval study in the Avon River (pers. comm.). This suggests that the factors responsible for the timing of the life­ history, and growth of larvae, were similar in both places. 126

5.3.2 Triplectides cephalotes (Trichoptera: Leptoceridae)

The size frequency distribution of larvae indicated the existence of five instars (Fig. 5.5). The identity of the first instar was confirmed by measurement of larvae taken from gravid females since females of this species are ovoviviparous, giving birth directly to

500 - 600 first ins tar larvae (Pendergrast & Cowley 1966).

40 2 3 N :; 576

n 5

O+---~L-~~--~~=U~--~~--~------~~~-- a 1.0 2.0 hind tibia length (mm)

Fig. ?5 Size frequency distribution of hind tibia lengths of Triplectides cephalotes larvae (all data combined September 1976 - March 1977).

Although most instarswere present in most months (Fig. 5.6), suggesting that larval recruitment and growth were not well synchronised, the general pattern of growth of the main component of the population could be followed. Very few first instar larvae were collected in field samples (and only in April 1977). Since larvae, rather than eggs, are released by the females, it is probable that a rapid moult into the second instar occurs to take advantage of the release of young at a relatively advanced stage. First instar larvae could be recruited to the population between late December and April. This could not occur at the onset of the adult flight period because time would be required for development to the first instar inside the adult females. After a period of rapid growth, most larvae (40 - 70%) spent the winter in the third instar (cf. H. amabilis, the fourth instar), and passed through the final two instars from late October to January. Pupae were found attached to macrophytes from early NoveffiQer to late March (Fig. 5.6), but adults were taken only in December and January (Appendices 5.3 and 5.4) • Although there were no adult collections taken in February, and adult T. cephalotes were not taken in March, it was probable that the 127

N 33 42 173 25 19 50 43 19 17 19 43 30 21 42 c:======:::. A p 2.5

E 2.0 5 E

3 0.5 2

O~A~--S---O----N---D-''-J~~F~~M~~A~~M~-J~--J~-A----S~--O~-N~--D-'-J~--F~-M-'~ 1976 1977 1978 mean size in dominant ins tar 0-'100%

Fig. 5.6 Life-history pattern of Triplectides cephalotes (September 1976 - March 1978) showing the percentage of larvae in each instar and periods of adult and pupal abundance. N number of larvae collected.

flight period extended from about the end of November to April (based on pupal evidence). MCFarlane (1977) noted a flight period from 23 November to 19 April at the Winchmore Irrigation Research Station on the Canterbury Plains, and Cowley (1978) collected adults from September to mid May around Auckland in the North Island.

Cowley (1978) found final instar larvae in a swimming pool in Auckland on 30 January, and suggested that the first instar larvae could not have been before mid December (presumably because there was no water in the pool before then). This indicates that, at least in warmer areas of New Zealand and at the inflated temperatures often found in swimming pools, the life-history can be short.

5.3.3 pycnocentrodes aureola (Trichoptera: Conoesucidae).

Limited life-history information was collected for P. aureola, a conoesucid caddisfly that lives on stony or rocky substrates, rather than macrophytes, in Lake Grasmere (and more commonly in streams (Cowley 1978».

Larval ins tars could not be distinguished on the basis of head width measurements (Fig. 5.7), but there was apparently only one cohort present during the year, and a univoltine life-history pattern (Fig. 5.8). 128

20 e 8UREOLA il n=266

o 0,5 1.0 1.5 head width (mm)

Fig. 5.7 Size frequency distributions of pycnocentrodes aureola (all data combined, March 1977 - March 1978).

n :: 1 17 37 9 30 14 30 44 13 77 A ------P ------w <.!)

I-« (j') 1 >- 0::: L 0 I- (hw. (j') mm) --- " " " " ,," " o E M J J A o N o J F M 1977 1978

Fig. 5.8 Diagrammatic representation of the life-history of pycnocentrodes aureola (March 1977 - March 1978). Vertical bars show the size ranges of larvae collected each month; mean sizes are joined. Presence of eggs (E), pupae (P), and adults (A) are also shown. n = number of larvae sampled. Dashed lines represent life-history stages that were presumed to be present but were not recorded.

Spherical egg masses characteristic of P. aureola were found amongst rocks, stones and debris in shallow water along the eastern shore of Lake Grasmere from mid November to early May, and pupae were recorded in November and December (but may have been present also in October, January and early February). Adults were taken in hand-net and light- trap collections from early November to late January, occasionally in 129 very large numbers (Appendices 5.3 and 5.4). Norrie (1969) collected adults throughout the year near Auckland, and Cowley (1978) recorded the presence of larvae and pupae at all times, although adults were more common in spring and summer. McFarlane (1977) noted a flight period from mid October to early May at Winchmore in the South Island. However, it seems that the flight period at Lake Grasmere was shorter, since adults were not collected on 27 October (1978), 2 March and 8 April (1977) (Appendices 5.3 and 5.4).

5.3.4 Oecetis unicolor (Trichoptera: Leptoceridae)

O. unicolor was most common in Lake Grasmere on sandy substrates and amongst low macrophytes, especially near the outlet stream. It was not collected often in samples from macrophytes (see Chapter III).

In Lake Grasmere, O. unicolor apparently has an annual life- history with a single cohort present in anyone year ( . 5.9). The pattern is virtually identical to that for P. aureola (see Fig. 5.8), and also for animals collected from Lake Pearson by Dr B.V. Timms (pers. comm.). No attempt was made to locate eggs, and pupae were not recorded.

n =5 9 5 2 36 2 .5 13 26 4 A ------W (!) p i5! ------(f) >- 5 a::: 0 I- 4 (f) .I 3 w I LL 2 t )-- " I ..-I I;' " 1

E -----_ ..... _--- _ ...... _------s a N 0 J F M A M J J A S a N D J F M 1976 1977 1978

Fig. 5.9 Diagrammatic representation of the life-history of Oecetis unicoi or (September 1976 - March 1978). Vertical bars show the range of instars present each month; dominant ins tars are joined. Presence of eggs (E), pupae (P), and adults (A) are also shown. n = number of larvae sampled. Dashed lines represent life-history stages that were presumed to be present but were not recorded. 130

Adults of O. unicolor were collected in December and January (Appendices 5.3 and 5.4), but since larvae were most common at the opposite end of the lake to that where most collections for adults were made, these collections are unlikely to be a reliable indication of the flight period. Cowley (1978) collected adults from October to mid April, but suggested that the flight period was longer than this because gravid females were taken in April. McFarlane (1977) recorded adults in light-trap collections from Winchmore between mid October and late March, and the seasonal abundance of larval instars in Lake Pearson (about 2.5 km from Lake Grasmere) during 1978 - 1979 indicated that adults emerged mainly in December, January, and February (B.V. Timms pers. comm.).

5.3.5 oxyethira albiceps, Paroxyethira hendersoni and P. tillyardi (Trichoptera: Hydroptilidae)

since specific separation of larvae of the first four instars of New Zealand hydroptilids is not yet possible (see Chapter VI), and there is little intraspecific size variation in final instar larvae of all three species, I did not measure larvae collected on a regular basis for life-history analyses. However, information on the presence and abundance of final instar larvae and pupae obtained in the quantitative sampling program (Appendix 2) and non-quantitative hand-net collections, and adults from hand-net and light-trap samples (Appendices 5.3 and 5.4), has enabled the elucidation of certain features of their life-histories.

O. albiceps was rare in quantitative samples from macrophytes but more common on stony substrates amongst filamentous algae at the southern end of the lake where final ins tar larvae occurred throughout the year. Adults were collected in six months of the year (January, March, May, October, November, and December) (Appendices 5.3 and 5.4), but it is likely that emergence can occur in almost any month (since pupae were recorded even in mid June) . Cowley (1978) noted the presence of all life-history stages throughout the year, and McFarlane (1977) collected most adults in light-traps at Winchmore between September and May, although some were collected during the remainder of the year.

Final instar larvae and pupae of P. hendersoni, the most common hydroptilid in Lake Grasmere, were collected in almost every month in which samples were taken (Appendices 2, 5.3 and 5.4), although pupae were not recorded in July. Most emergence seemed to occur between 131

October and March or April (Appendix 5.3). McFarlane (1977) noted a flight period from early November to mid May at Winchmore, and Cowley (1978) stated that adults could be found at any time of year.

Final ins tar larvae of P. tillyardi were recorded in all collections, but pupae were found only in December and January (despite' intensive at other times), and adults mainly in December and January but occasionally as late as March (Appendices 2, 5.3 and 5.4). This species had the shortest flight period of the three hydroptilids (and, in fact, of all insects studied) in Lake Grasmere and, apparently, a well synchronised emergence.

5.3.6 Nymphula nitens (Lepidoptera: pyralidae)

I was not able to ',determine the number of larval instars in this species by or inspection of raw size frequency data (Fig. 5.10), but application of Dyar's rule, which states that there is a geometric increase in the size of a sclerotised structure with each ins tar in insects (Crosby 1973), indicated that there were probably nine larval instars (indicated on Fig. 5.10). The head width increased by a factor of 1.28-1.29 at each moult. The size of the first instar larvae

(hw = 0.20 rum) was established from larvae hatched from eggs found on Myriophyllum propinquum.

7 V 40

N. NIT ENS 6 n = 461 V n

O+-____~~~~==~-+ ______._------._------_+-L-----W~~ o 0.5 1.0 1.5 head width (mm)

Fig. 5.10 Size frequency distributions of Nymphula nitens (all data combined, July 1976 - March 1978). 132

N. nitens has a univoltine life-history pattern in Lake Grasmere (Figs. 5.11 and 5.12). First instar larvae were collected in January, and grew rapidly until late March when most were probably.in instars 5 to 7. There was very little growth in winter when mid-lake water temperatures were below 11 or l2°e (see Fig. 2.3). Growth resumed in October and continued through to emergence in December and January.

[J o 1.5 CJ IT o [] CJ o o 0 o CJ o o o CJ o U o Cl El.O o :; 0 ....,E <; y 0 CJ c:::J 0 CJ o o CJ o o "0 CJ o CI1 CJ lJ o Q) .r:. 0.5 CJ

CJ CJ 'V n=19 19 22 29 10 10 16 19 17 12 8 11 6 12 257 OL-__~~------~~~-- ______.- ______3-7 2-9 2-11 2-12 20-1 2-3 8-4 10-5 20-6 13-718-8 3 -10 21-11 26-12 22-3 1976 1977 1978 NytllPHULA NIT ENS

Fig. 5.11 Size frequency distribution of Nymphula nitens larvae collected from Lake Grasmere between July 1976 and March 1978. n = number of larvae sampled.

The mean sizes of N.· ni tens larvae were greater in 1977 - 1978 than at corresponding times the previous year (Fig. 5.11), suggesting more rapid growth in 1977 1978. This may have been associated with higher water temperatures (1.5- 5.0o e higher from August 1977 to February 1978 than in the same period of 1976 -1977, Fig. 2.3). Also, numbers of larvae collected (with similar sampling effort) from the lake in early 1978 were about 25 times (1977) and ten times (1976) greater than in the previous two years (Fig. 5.11). In March 1978, larvae were found on their preferred food plant (M. propinquum) , and in large numbers also on Elodea canadensis. The occurrence of many larval cases constructed of E. canadensis leaves bound by silk threads, and observations of partly defoliated macrophyte stems, were striking evidence of the impact of hiqh larval densities. 133

E .§ t-- ~r~976'7 ..c ..- iJ 'j t~r , ,­ , "0 I rd QI ..c -.1 JASON D J FMAMJJ ASO NDJ FM 1976 1977 1978 NYMPHULA NITENS

Fig. 5.12 Life-history pattern of Nymphula nitens (July 1976 - March 1978) showing mean size, and range of sizes of larvae in each month sampled. The dashed line represents the growth curve of 1976 - 1977 superimposed on that for 1977 - 1978.

5.3.7 Xanthocnemis zealandica (Odonata: Coenagrionidae)

Deacon (1979) found that X. zealandica at Lake Sarah (about 1 km from Lake Grasmere) had a three-year life-cycle, but that some larvae could complete development in two years (cf. a lowland population where a two-year life-history was the rule). Part of my study period (September 1976 - March 1977) overlapped with that of Deacon (1979), and his more complete data facilitates the interpretation of the life-history of X. zealandica at Lake Grasmere.

Fig. 5.13 shows the size frequency ,distribution of larvae with cohorts (i.e., larvae of the same year-class) numbered 1- 4, and approximate growth curves of component cohorts. The striking feature of Fig. 5.13 is the poor representation of cohorts 3 and 4, whose absence is unlikely to be due to sampling inadequacy since large numbers of young larvae were taken in September 1980 (Fig. 5.13). Instead, their low representation may indicate poor recruitment of these year classes, which could have been the result of anyone (or more) of a number of factors (e.g., poor parental emergence, mating success, oviposition success, hatching success or survival of small larvae).

Eggs are laid during the summer inside the stems and leaves of floating plant material or rooted macrophytes that reach near the water surface. The first winter is spent as early instar larvae (e.g., cohort 4, Fig. 5.13), but by the following March some larvae have reached

1.5 rom hw. Little growth occurs during the second winter (e.g., cohort 134 emergence periods (from Deacon 1979) to late Mar.

/' /' /'

3 / -- ..... "'" 20% 4_ '-----' -III - - - 69 24 20 9 70 43 16 17. 31 487 h ·····-5- M I A I M I J I J I A 5 I 0 N I 0 1977 r 1980

Fig. 5.13· Size frequency distribution of xanthocnemis zealandica from Lake Grasmere (September 1976 - December 1977), with an additional collection from September 1980). Lines indicate approximate growth curves of component cohorts. The dashed lines are based on data from Deacon (1979) in the absence of larval representation in samples from Lake Grasmere. Also shown are the presumed emergence periods (from Deacon 1979) and numbers of larvae collected. Arrows at left of September 1980 size frequency distribution indicate approximate divisions between cohorts.

3, Fig. 5.13), but early in the next year some larvae reach the F-l and final instars (e.g., cohort 2 in March 1977, Fig. 5.13). However, most larvae spend the third winter (when, once again, there is little growth) as F to F-3 instar larvae (e.g., cohort 2 in winter 1977, Fig. 5.13) and emergence occurs the following summer. The pattern of larval growth, in summary, is three winters of very little growth, each followed by a period of rapid growth during the summer, with the third culminating in emergence.

Adults were collected from early December 1976 to 8 April 1977, and again on 9 October 1978 (Appendix 5.7). Deacon (1979) found that emergence could start on different dates at various sites, but only when the water temperature had risen to 10 - 12°C. At Lake Sarah (lake shore site), the emergence period ranged from 126 (1976 - 1977) to 147 (1977- 1978) days, beginning between 14 and 22 November in 1976, and 6 and 14 November in 1977 (Deacon 1979). The termination of emergence was between 16 and 20 March in 1977, and 26 March and 2 April in 1978. These dates may apply equally well to the emergence of X. zealandica from Lake Grasmere since the temperature regimes of the two lakes are 135 usually similar (stout 1969a,Deacon 1979, cf. present study). Deacon (1979) noted a temperature range from 4°C to 20°C in Lake Sarah between early July 1977 and late April 1978, which is the same as that recorded in the study (Fig. 2.3) from Grasmere. The earlier warming of Lake Grasmere in 1977- 1978 (cf. 1976 1977), and presumably also Lake Sarah, is consistent with the earlier emergence of X. zealandica in 1977 - 1978 (as noted by Deacon (1979». The collection of nine adults on 9 October 1978 from the southern end of Lake Grasmere, that emergence in that year was even earlier than in the two preceding years. I I

I I

I I

I I

I I

I I

I I 137

CHAPTER VI

TAXONOMY OF THE NEW ZEALAND HYDROPTILIDAE (TRICHOPTERA) AND CHIRONOMIDAE (DIPTERA)

6.1 INTRODUCTION

Invertebrates collected during the course of this study that could not be identified readily to the specific level, in particular, adult and larval Chironomidae and the larva of a species of Paroxyethira (Hydroptilidae), stimulated an interest in the taxonomy of these groups. I intended, initially, to identify specimens collected from Lake Grasmere (both larvae and pupae from aquatic habitats and adults collected in light-traps or hand-nets), but this work developed into a more intensive study of the taxonomy of hydroptilids and chironomids and the construction of keys to the New Zealand species (see also Stark in press) •

This chapter presents keys to the larval Hydroptilidae and to the larval and adult male Chironomidae of New Zealand constructed from the literature and from my own taxonomic investigations, and discusses taxonomic problems concerning these groups. Distributional records and habitat notes are included where possible, since these can be useful aids to taxonomic studies. Descriptions of the larva of Paroxyethira tillyardi and the adult males of two unnamed chironomids (subfamily Orthocladiinae) are given together with notes and figures of chironomid larvae and pupae from Lake Grasmere.

6.2 TAXONOMY OF THE LARVAE OF NEW ZEALAND HYDROPTILIDAE

Of the six species of Hydroptilidae described from New Zealand, one is in the widespread genus Oxyethira Eaton 1873 and five in the endemic paroxyethira Mosely 1924. Adequate descriptions of adults were given by Mosely (1924) and Leader (1972), and.a key to adults by Leader (1972) • The larvae of only two species (0. albiceps and P. hendersoni) have been described by Cowley (1978) who referred also to previous work on the family in New Zealand.

The following key to larvae and pupae of the New Zealand species of hydroptilids includes all of the described species, although it is not yet possible to distinguish between P. eatoni, P. hintoni and P. kimminsi. Pharate adUlts (i.e., late pupae that show the 138 characteristic adult genitalia) must be examined to enable specific identification of these species (see key in Leader 1972).

Efforts to differentiate the species further have been frustrated by a paucity of larval material of P. kimminsi (especially) and P. hintoni. The marked similarity between adults of P. kimminsi Leader 1972 and P. eatoni Mosely 1924 deserves further consideration when more material is available. In addition to the characters used in the key, the morphology of larval prosternal plates was examined and four species-groups were evident (Fig. 6.1). However, no further differentiation of species was possible.

A

Fig. 6.1 f. hendersoni Larval pro sternal plates and cases of New Zealand c Hydroptilidae.

o C_:t P.~ p . b.in.1Q.Dj p.kimminsi

I.--.-.J o 50 0----10 J.rn mm

There has been some suggestion that more than one species of Oxyethira is present in New Zealand (Michaelis 1977, A.G. McFarlane pers. corom.). This belief appears to be based on the considerable size variation among larvae, pupae and adults and minor differences in larval case morphology of specimens from some areas. I do not consider that, at present, there is evidence sufficient to warrant the erection of another species. Towns (1976) also found no evidence to support the existence of more than one species. 139

6.2.1 ~L-~~~~~~~~~~~~ ____~~o~f~N~e_w __ z~e~a~l~a~n~d (from Stark in press)

1. With a case ...... •.< •••••••••••••• < •••••••••••••••••••• 2 Without a case; either ins tars 1-4 (of any species), or 5th instar larvae which have lost their cases .•..••...... ••...... 5

2. Case the shape of a flask or axehead (Fig. 6.1A) ..••.....•...... • ...... Oxyethira albiceps (McLachlan) [Widely distributed in freshwaters from sea-level to about 900 m. All life history stages exhibit considerable variation in size and may be found throughout the year. (Cowley 1978)] Case approximately rectangular. (Fig. 6.1B-D) ..•.. Paroxyethira 3

3. Case with spine-like projections which interrupt its otherwise smooth outline ....•....•.•...... •.••...... ••...... •... 4 Case without projections (Fig. 6.1D) •.••...... P. eatoni Mosely P. kimminsi Leader P. hintoni Leader [These three species are not well known and as yet their larvae cannot be separated. P. eatoni inhabits small ponds, seeps, tarns and the quieter stretches of streams and often is associated with filamentous algae and diatoms. P. kimminsi is known only from quieter of streams in the Waitakere Ranges near Auckland. P. hintoni is known from mountain streams above 600 m. (Leader 1972)]

4. Case with two horizontal projections at each end (Fig. 6.1B) .... P. hendersoni Mosely [Larvae occur widely in a variety of freshwater habitats from sea level to at least 1320 m. Case shape and size can be highly variable. (Cowley 1978, Leader 1970, Leader 1972)] Case with up to four lateral and two dorsal projections; often darkly pigmented, sometimes black (Fig. 6.1C) •.. P. tillyardi Mosely [Known only from lakes where larvae are found on algal-coated macrophytes and stones. Larval cases vary considerably in shape and in number and development of spines.]

5. Prothorax narrower than head; abdomen tapering gradually posteriorly and possessing long setae; anal appendages long and slender Instars 1-4, all species [At present it is not possible to separate instars of the six (Leader 1972)] 140

Pro thorax as broad as or broader than head; abdomen large and swollen, usually laterally flattened •...... Final ins tar larvae [P. hendersoni and P. tillyardi possess a complex spine on the median of the ventral prolongation of the protarsus but this spine is absent from O. albiceps. In addition, the tarsal claw of the hind limb of P. hendersoni is about 20 times as long as its basal width whereas it is only 10 times as long in P. tillyardi. The condition in the other Paroxyethira species is not known.]

6.2.2 Distribution of the New Zealand Species of Hydropti1idae

o. albiceps was considered by Leader (1972) to be by far the most abundant and widespread of the New Zealand Hydroptilidae, being recorded from Dargaville in the North Island to Dunedin in the South Island (Fig. 6.2A), from lakes and streams between sea level and about 900 m a.s.1. Cowley (1978) suggested that, in most freshwater habitats, its distribution was related more to food availability (probably filamentous algae) than to current speed or substrate. O. albiceps may be found in virtually any freshwater habitat with sufficient filamentous algae and this is reflected in its wide distribution within New Zealand and its presence on Chatham Is., Auckland Is., Campbell Is., Snares Is. and Antipodes Is. (Wise 1972, 1973, 1978).

P. hendersoni also is found throughout New Zealand (Fig. 6.2B) in lowland lakes and ponds, on aquatic macrophytes at the edges of large rivers (Leader 1972) and in the outlet stream of L. Turbott, Auckland Is. (Wise 1972). I have found this species in a wide variety of fresh water habitats (viz., most freshwater bodies containing macrophytes, filamentous algae and/or diatoms) and have records of adults from 1320 m a.s.l. (Lake Sylvester, N.W. Nelson, February 1966, A.G. McFarlane colI.).

P. tillyardi*, however, has been found only in lakes, where the larvae occur on filamentous algal or diatom encrusted macrophytes or stones. Leader (1972) stated that P. tillyardi appeared to be a highly localised insect because, despite intensive collecting, it was known (as an adult) from only two places (Lake Ianthe on the West Coast of

* All references to P. tillyardi in Leader (1970) are actually P. hendersoni since Leader misassociated P. tillyardi adults with P. hendersoni larvae from Lake Ianthe. I have found the larvae of both species in samples collected by Dr K. Deacon from this lake. 141

f

" e henders~ni

Fig. 6.2 Distribution of (A) Oxyethira albiceps and (B) Paroxyethira hendersoni. 142

• e ealonj x E. hinlQnj ( _____ 0 E. kimminlii ______

Fig. 6.3 Distribution records of (A) paroxyethira tillyardi and (B) P. eatoni, P. hintoni and P. ki~insi. 143 the South Island and the Tarawera region, North Island). P. tillyardi has a very short flight period (see Chapter V) and would seem to be comparatively rare if its presence was determined solely on the basis of adult records. Larvae are present all year round, and I have found them in collections from at least 14 different lakes throughout New Zealand (Fig. 6.3A). The geographical range of P. til1yardi appears to be limited by the presence of suitable larval habitats, hence, for example, its absence from the Canterbury Plains.

P. eatoni, P. hintoni, and P. kimminsi are poorly known as larvae and, as yet, cannot be distinguished from one another in the absence of pharate adult or adult associations. P. kimminsi is known only from its type locality in the waitakere Ranges, Auckland (Cascades Stream). Leader (1972), who described the adults, also recorded the presence of larvae and pupae associated with filamentous algae and diatoms on the sides of stones in quieter parts of the stream. Maximum abundances of immature stages occurred in summer. Towns (1976) also recorded its presence in the same area (Fig. 6.3B). P. hintoni was described by Leader (1972) from adults collected near the TePopo Stream (Mount Egmont, altitude 700 m) and he also collected adults and recorded immature stages in a small stream flowing into the north side of Lake Tekapo (altitude 600 m+) (Fig. 6.3B). I have also examined specimens from the Edwards River in the Arthur's Pass National Park (760 m a.s.l., 20 January 1939, A.G. McFarlane colI.) (Fig 6.3B). Thus, this species seems to be restricted to streams above about 600 m a.s.l. P. eatoni is more widely distributed (Fig. 6.3B). The adult male was described by Mosely (1924) from the Tekapo River in the Mackenzie County, South Island, and the female by Leader (1972) who did not mention a type locality. Dr J.P. Leader (pers. comma 22 March 1977) has found larvae only at one place - 'in a stream running underneath a railway bridge near Inangahua' (West Coast, South Island). Adults of this species are 'commonly caught in light-traps around the North Island lakes' (Dr J.P. Leader pers. comm. 22 March 1977). I have examined adult material from Lake Moeraki (3 February 1965, A.G. McFarlane), Lake Ohau (12 March 1966, A.G. McFarlane), Lake Rotoiti (South Island) (22 November 1978, B.V. Timms) and larvae and pupae from tarns and seeps near Lake Sylvester (1320 m a.s.l., A.L.P. colI. in A.G. McFarlane colIn), 'Kaituna', Banks Peninsula (13 September 1964, J.G. Penniket) and a small pond in the bed of the puhipuhi River near Kaikoura (6 November 1978, J.D.S.). It seems likely, since larvae have not been collected 144 from lakes, that the preferred habitat of larvae is small ponds, seeps, tarns or slow-flowing backwaters of streams and rivers in association with filamentous algae and diatoms as at the Puhipuhi River.

6.2.3 Description of the Larva of Paroxyethira tillyardi (Figs 6.4, 6.5e,f) Paroxyethira tillyardi Mosely 1924.

Mosely 1924, Transactions of the New Zealand Institute 55: 670-673. LARVA (final instar): larval length variable 2.2 - 3.6 mm; case length 2 • 7 - 4. 0 mm, depth O. 9 - 1. 0 mm. CASE. The case of P. tillyardi varies greatly in size, colour and the number (0 - 4 per side) of the lateral and size of the dorsal projections (Fig.6.5e,f). It can be up to 4 'mm long (usually 3.6-3.7 mm) and 1.0 mm deep. The distance between the dorsal processes varies between 1.2 and 1.6 mm. The case is constructed of medium to dark brown-black semi-transparent secretion and is usually lighter in colour towards each end. The two ends of the case are identical. HEAD (Fig. 6.4b). Length 0.25 mm, width 0.19 mm, L/W = 1.32. Only slightly elongate, oval, almost parallel-sided. Pigmentation dark brown-black with pale brown around large eyes. Frontoclypeal apotome wider across lower half than upper. Antennae prominent with lateral bristle at least 1.5x antennal length. Pregula with narrow lateral arms; postgula a crescent-shaped plate (Fig. 6.4c). Mandibles asymmetrical, with two outer basal bristles. Left mandible with a well-formed inner brush of hairs (Fig. 6.4d). Facial hairs very long, some 2X length of head capsule. THORAX. All three notal plates well formed. Mesonotum and metanotum similar in length and breadth (0.16 - 0.18 x 0.23 - 0.26 mm) (L x W) and clearly wider than pronotum. Prothorax. Length 0.18 mm, width 0.21 mm, L/W 0.86. Pronotum with straight anterior edge and evenly curved anterolateral angles. Pigmentation mid-dark brown, lighter anteriorly and laterally. Anterior edge of pronotum bearing (from midline laterally), 1 short, 1 long, 2 short, I long and 5 short setae, and, on the dorsal surface 2 pairs of medium length setae. Prosternum (Fig. 6.1) with a single T-shaped oral sternite. Meso- and metathorax. Well-formed; pigmentation mid-dark brown, darker posteriorly. Mesosternal region with small, elongate trapezoid sclerite on mid ventral axis and 2 irregular, transversely elongate anal plates. Metasternum with weakly sclerotised Y-shaped oral 145

c

O.lmm

d

Fig. 6.4 Paroxyethira tillyardi: a. antenna, b. head, c. gular d. mandibles, e. claw of hind limb (e' = claw of hind limb of P. hendersoni). All scale bars 0.05 rom unless indicated otherwise.

sternite and 2 very narrow, transversely elongate anal sternites. Legs. Forelegs short, sturdy, Middle and hind legs longer and thinner; pretarsal claws relatively stout (cf. P. hendersoni Fig. 6.4e, eO). Forelegs uniform pale brown-gold. Tibia with peg-like projection bearing 2 distal setae - one blade-like and one stout - and a complex comb-like spine on the inner base. Ventral edge of femur with several prominent setae arising from slight projections. 146

b

Paroxyethira tillyardi

Fig. 6.5 a. Oxyethira albiceps cased larva; b. Paroxyethira eatoni pupal case; c. P. hendersoni cased larva; d. P. hendersoni case; e. P. tillyardi cased larva; f. P. tillyardi case. (Scale bars = 1 mm)

ABDOMEN lime green (bright yellow rapidly leaching to white when preserved in 70% alcohol), without gills. Segments 1 and 9 small, clearly defined. Segments 2 - 8 large, segments 4 -6 largest. Segment 9 with a medium-sized, oval tergite bearing 4 pairs of hairs, the middle pair very long. Segments· 3 - 7 with a minute chitinous ring mid-dorsally. Anal appendages shortened. Segment 10 short, compact. Dorsal supporting plate with 2 setae, one on surface the other arising from a notch in distal edge. Claw with a short, strong primary hook and 4 small auxiliary hooks, the second pair very slender.

SPECIMEN LOCALITIES*

South Island. NN - Boulder L. 27 October 1963 V.M. stout, MC - L. Coleridge 14 January 1968 A.G. McFarlane; L. Grasmere 14 April 1976 - 21 October 1980; WD L. Brunner 10 January 1965 V.M. stout; MK - L. Ohau 19 February 1976 V.M. Stout; FD - L. Te Anau 30 September 1976 V.M. Stout; L. Lochie January 1962 G.A. Knox.

Code system for specimen localities follows that recommended by Walker & Crosby (1979).

DIAGNOSIS The presence of a complex comb-like spine on the inner base of the fore-tibia separates larvae of paroxyethira from Oxyethira. 147

P. tillyardi differs from other Paroxyethira species in the form of the prosternal plates, the darker pigmentation (especially of the head and thorax), and the more robust limbs. The larval case is very different from those of the other species (Fig. 6.5).

6.3 TAXONOMY OF NEW ZEALAND CHIRONOMIDAE

Few larvae of New Zealand Chironomidae have been described (Brundin 1966; Forsyth 1971, 1975b, 1979; Sublette & wirth 1980) and consequently identification, even to the generic or tribal level, has been a problem. On the other hand, adult .taxonomy is reasonably well advanced as a result of the work of Pagast (1947), Freeman (1959, 1961), Brundin (1966), Forsyth (1971, 1975b) and Sublette & wirth (1980).

The following keys to larvae and adult males of freshwater chironomids known from the main islands of New Zealand include all described species of the subfamilies Tanypodinae, Podonominae, Diamesinae and Chironominae (and four undescribed Chironominae). Orthocladiinae, Telmatogetoninae (marine) and chironomids known only from New Zealand's subantarctic islands (see Sublette & Wirth 1980) are not included. Orthocladiinae systematics is confused by various conflicting systems of nomenclature, many synonymies (see, for example, Hamilton, Saether & Oliver 1969), and the proliferation of very narrow generic diagnoses (as evidenced by, for example, eight monotypic genera erected by Sublette & Wirth (1980) for orthoclads from New Zealand's Subantarctic islands). The existence of many undescribed species makes it inadvisable to construct keys to New Zealand Orthocladiinae.

The following keys to the subfamilies and species of larval Chironomidae, complete with introduction, are based on the literature and my own taxonomic investigations. These keys are included in the compilation of keys to the freshwater insects of New Zealand by Winterbourn and Gregson (in press) and are the most comprehensive keys to New Zealand chironomids constructed to date

The identification of chironomid larvae is not easy since taxonomic knowledge of larvae lags behind that of adults. Therefore, it is often necessary to establish the link between adult and larval stages of a species to enable specific identification. 'The importance of life history information (in the broadest sense, the ecology of a species) in this regard should not be underestimated. Behavioural features, microhabitat preferences and data on abundance can all be 148 useful in associating life-history stages. Often a clue to larval identity can be obtained by association with adults collected in the region of the larval habitat although there is always the danger of misassociation. Pupae can be especially useful because they develop adult characteristics (e.g., genitalia) and it may then be possible to associate adults, pupae and larvae through field collections. The best method, and the one least subject to misinterpretation, is that of rearing individually isolated larvae, through the pupal stage, to emergence as the adult. One then has larval head capsules, pupal exuviae and adults for examination. If due care is taken to duplicate environmental conditions (especially temperature, current speed, substrate and food supply) rearing is not difficult for many species (e.g., several species reared by Forsyth (1971) took 12-25 days at 20-25°C to grow from egg to adult). Identification of chironomid larvae relies on features of both the head and body and usually requires the mounting of specimens on slides. The following procedure is recommended: Colours of specimens should be noted before they are killed and stored in 70% alcohol. To mount a larva on a slide, the body should be separated from the head and mounted on its side whereas the head should be placed ventra1 side up. It is often best to boil the head in 5-10% KOH (10 min or less) to digest away muscle tissue prior to mounting on the slide. A good mounting medium is lactophenol-PVA. Sometimes, temporary mounts in water are useful for examination of the fine structures of mouthparts. The nature of the taxonomic characters used in the keys will become evident on referring to the figures; anatomical terminology used follows that of Mason (1973).

I would like to emphasise that the keys must be used with caution as the New Zealand fauna is relatively poorly known. Especially within the subfamily Orthocladiinae, there are many undescribed species, a number of which are extremely similar as adults and, as yet, not separated as larvae. Overseas keys (Bryce & Hobart 1972, Mason 1973, Martin 1974, 1975, Oliver, McClymont & Roussel 1978) also may be useful for identifying larvae at the generic level.

Nomenclature used below is that employed by most contemporary European workers (e.g., Brundin 1966, Fittkau 1962). As such; it is in line with recent work in Australia (J. Martin,' pers. comm.) and current trends in North America (Hamilton, Saether & Oliver 1969, Saether 1977). The European classification employs smaller genera than those traditionally used by most North American taxonomists and past New Zealand workers (Forsyth 1971, Freeman 1959). 149

As the distributions and habitat requirements of most species are poorly known, few annotated notes on biology are appended.

6.3.1 to Larval Chironomidae of New Zealand

[Subfamilies or tribes marked with a single asterisk have not been recorded from New Zealand. **Orthocladiinae, Clunioninae and Telmatogetoninae are poorly known in this country as larvae and are not keyed further.]

Key to subfamilies 1. Head capsule with fork-shaped lingua; antennae retractile into sheaths embedded in head (Fig. 6.6a,c) ••••••... Tanypodinae, p.150 No fork-shaped lingua; antennae not retractile •..•••••••.....•... 2

2. Premandibles absent ••••...... ••••.•....•.•.•••.•••.....••••..•... 3 Premandibles present (Fig. 6.6d) •.••••.•...••..••.•.••...•.•••.•• 4

3. Posterior procerci (= preanal papillae) 5 to 10 times as long as wide; antennae 4 br 5 segmented, third segment may be annulated; hypopharynx with ventral lamellae projecting forward •...... •....• Podonominae, p.15l Posterior procerci lacking; antennae 3-segmented, not annulated; labial plate without teeth, anterior margin nearly straight; body heterogeneously sclerotised, partly covered with plates of differing forms and bearing strongly developed setae ...•.••...... • . . • . . . . • • • • • ...... • . . • • • • • . • ...... • • . . . . • • • . • . . • . • •. Aphroteniinae *

4. Paralabial plates with striations (Fig. 6.6e) (exception: Harrisius pallidus, paralabial plates indistinct) .•.•...••.••.... • • • ...... • • • • ...... • • . . . . • • • . • . • • • . . • • • • • • . • . . . • . •• Chironominae, p. 153 Paralabial plates, if present, without striations (Fig. 6.6d) .... 5

5. Third antenna 1 segment with annulations OR head, in dorsal or ventral view, tapering towards the front (i.e., tending toward trapezoid shape), occipital margin with distinct, deep-black neck (Fig. 6.7a), head colour either dark reddish-brown or light yellow, head capsule often with numerous long setae ..•.•••••• Diamesinae, p.152 Not as above; third antennal segment never annulated ..••.•••....• 6

6. Freshwater species (some terrestrial or semiterrestrial); labial plate variable, usually convex anteriorly, its central three-

quarters always with teeth (Fig. 6.6d) •..•••••. Orthocladiinae~* p. 157 Generally marine (intertidal) .•... Clunioninae & Telmatogetoninae** 150

Key to larval Tanypodinae

1. Paralabial combs (Fig. 6.6c) absent; abdominal segments slender, without hair fringe; anal gills slender ...••. Tribe Pentaneurini, 2 Paralabial combs, or a row of free chitin points present;

abdominal segments broad, usually with hair fringe ..••.... ~ ...•.. 3

2. Maxillary palp with more than one basal segment (Fig. 6.6f); lingua with 5 teeth ...... ••....••..• Ablabesmyia mala (Hutton) [Found in lakes, ponds and streams. Sublette & Wirth 1980, Hutton 1902, Freeman 1959] Maxillary palp with a single basal segment; lingua with 5 teeth · • . . • . . • . . • . . . . • ...... • . . . . • • . . • • . . • • . . • • • . . . • . .. other Pentaneur ini [One described species, pentaneura harrisi Freeman, and at least one undescribed species. There'is some doubt concerning the generic placement of Australasian pentaneura (J. Martin pers comm.). Found in streams and lakes. Freeman 1959, Forsyth 1971]

3. Antennae at least half as long as head; a row of free chitin points present in place of paralabial comb; lingua with 6-7 teeth ...... •...••....••...... •...•....•...•• Tribe Coelotanypodini * Antennae at most one third as long as head; paralabial combs present ...••...... •••..•.••.•.•.•....••..•..•.•..••...•••...•.••. 4

4. Mandible with thick, bulging basal portion; 6 anal gills · . . • . . . . • • . . . . • . • • • . . • • • . . . . . • • . • . • • . . . . • . . • • . • . .. Tribe Tanypodini * Mandible not as above (Fig. 6.6c); 4 anal gills .••.....•.....•••. 5

5. Lingua with 4 yellow teeth of equal length, OR lingua with 5 black teeth; super lingua scale-like with toothed edge • ..••••••..••.....•..••••••.•.••..•• Tribe Macropelopiini (in part) * Lingua with 5 reddish-yellow or brownish-black teeth; superlingua two-pointed •...•....••..••••.....•....•.•.•.....•.••...... ••.•••• 6

6. Mandible with large two-pointed tooth; labial plate with long pustule-like appendages latero-basally; paralabial combs each with 13 teeth ..•••.•..•••..••.•..•••••..••....•.••.. Tribe Anatopyniini * Mandible with two small teeth close together; no pustule-like appendages; paralabial combs each with, at most, 9 teeth; toothed margin of lingua concave or straight .•...•••...••..•...... •••.. · • • • . . . . • . • • . • • • • . • . • . • .• . . • . . • . . • • • . Tribe Macropelopiini (in part) [Larvae keying here belong to the genera Macropelopia, Apsectrotanypus or Gressittius. Ten species have been described as adults from New Zealand: Gressittius antarcticus (Hudson), Macropelopia apicincta 151

(Freeman), M. languidus (Freeman), M. debilis (Hutton), M. quinquepunctata (Freeman), M. flavipes (Freeman), M. apicinella (Freeman), M. umbrosa (Freeman) and the quadricincta Freeman and cana Freeman which may belong in Apsectrotanypus (J. Martin pers. comm.). Larvae of few of the above species have been recognised and genera and species are separated easily only as adults. Members of the tribe occur in many freshwater habitats. Forsyth 1971, Freeman 1959, Hutton 1902, Sublette & wirth 1980]

Key to larval Podonominae [The diagnoses of New Zealand podonomid larvae leave much to be desired. Tentative larval identifications should be checked by examination of pupae or adult males.]

1. Posterior procerci (= preanal papillae) uniformly pigmented •••... • • • • • • • • • . • • • • • • • • • • • • • • • • • • . . • • . • • • • • . . . • • • • • •. Tribe Podonomin i, 2 Posterior black basally, hyaline transparent) distally ••.••.••.•...... •.••....•...•..•.•..•. ' Tribe Boreochlini *

2. Antennae comparatively short and stout, third segment annulated in most New Zealand species; middle tooth of labial plate considerably broader and longer than the first of 7 laterals (Fig. 6.6g); mandible with an apical group of 7 dark teeth (F ig. 6. 6h) ••.•.•..•..•.•••...... •.•.•••••••••••••• Parochl us spp. [Ten species have been described: P. conjungens Brundin, P. aotearoae Brundin, P. spinosus Brundin, P. maorii Brundin, P. ohakunensis (Freeman), P. carinatus Brundin, P. pauperatus Brundin, P. novaezelandiae Brundin, P. longicornis Brundin and P. glacialis Brundin. Specific determination is possible only by examination of pupal material although P. conjungens and P. glacialis can be identified as adult males. Common in mountain streams. Brundin 1966] Antennae short and stout, third segment never annulated; middle tooth of labial small, hardly broader or longer than the first of 7 or 8 laterals; head often broad and triangular but may be slender and parallel-sided . . . . • . . . • . • . • • . • . . • . • • • . • • • . • • • • • • • • • • • . • • • • . . . • • . • . •• Podonomus spp. [Three species have been described: P. parochloides Brundin, P. waikukupae Brundin and P. pygmaeus Brundin. Adult males are preferred for determination. Found in mountain streams. Brundin 1966] 152

[The larva of Zelandochlus latipalpis Brundin (recorded from Franz

Josef and Fox Glaciers) was described by Dumbleton (1973) I however, insufficient detail was given for this specieq to be included in the key. Larvae of Podochlus spp. are not known. Four species have been described: P. grandis Brundin, P. stouti Brundin, P. cockaynei Brundin and P. knoxi Brundin. Specific identification is possible by examination of pupae or adult males. Podochlus larvae probably inhabit mountain streams. Brundin 1966]

Key to larval Diamesinae

1. Third antennal segment annulated .•....••..•.•.•...•••...••..•.•.. 2 Third antennal segment not annulated •...•...... •....••.... 3

2. Dorsal surface of head with numerous large protuberances •...... • •. . . • . •• . . •• . . . .• • . . • • . • • • . • . • • . .• . ••. • . •• Tribe Boreoheptagyini* Dorsal surface of head without such protuberances ••••.•..•.••.•.• •...•••.. ...•••• .•••... •. . -.•.•••••.•••...... Tribe Diamesini*

3. Para labial plates well developed, extending beyond the labial plate by at least one half of the width of the labial plate. Paralabials with distinct beard of large, black hairs, or, hairs absent and central portion of labial plate with narrow concavity between two median teeth .••...... •..•..•..••... Tribe Prodiamesini* Paralabials not developed as above .••...•••...•••••••.•••..••...• 4

4. Anterior margin of labial plate virtually straight, middle one third of plate (at least) without teeth; premandibles not well developed and ending in a single blade •••..•••.• Tribe Protanypini* Labial plate distinctly convex (Fig. 6.7b,c); premandibles well developed and ending in more than one blade .•....••••....•....•.. 5

5. Antennae 4-segmented, basal segment more than twice the length of segments 2, 3 and 4 together; head yellow and body light green; long posterior prolegs; labial plate light yellow, median tooth nearly one half plate width and smoothly rounded, flanked by 7 darker laterals (Fig. 6.7c) Tribe Lobodiamesini [This tribe is monotypic, containing one species (Lobodiamesa campbelli Pagast) which is found characteristically in small, slow-flowing mountain streams. Brundin 1966, Pagast 1947] Antennae 5-segmented ...•••..•.•••••.....•.•...... • Tribe Heptagyini [Five species of Maoridiamesa belong in this tribe. They have 153

very dark, conspicuously triangular heads with black occipital margins produced into pronounced necks with two ventral, posteriorly directed, projections and dorsolateral incisions (Fig. 6.7a) labial plate with 15 teeth, median tooth broad, second laterals small, third laterals very large (Fig. 6.7b). Larvae of the five species, M. harrisi Pagast, M. intermedia Brundin, M. stouti Brundin, M. glacialis Brundin and M. insularis Brundin (Campbell Island) inhabit mountain streams and some lowland rivers, but only adult males or pupae can be identified easily. Brundin 1966, Pagast 1947]

Key to larval Chironominae

[The larvae of Tanytarsus albanyensis Forsyth, 1971 and Ophryophorus ramiferus Freeman 1959 are not known.]

1. Antennae arise from prominent tubercles (prominences), as long as wide or longer; first antennal segment long and curved; striated paralabial plates often nearly four times as wide as long and nearly touching in the midline (Fig. 6.7r) ....•...•.. Tribe Tanytarsini, 3 Antennal tubercles much wider than long, first segment not long and curved; striated paralabial plates usually (but not always) more fan-shaped (Fig. 6. 7k) .....•...... •..... 2

2. Paralabial plates nearly touching in the midline, about four times wider than long ..•...... •..•...•...... Tribe Pseudochironomini [One species, Riethia zeylandica Freeman has been described from New Zealand. The larva is unknown. Freeman 1959] Paralabial plates distinctly separated (or indistinct in Harrisius (Fig. 6. 7i)) •...... •...... Tribe Chironomini, 7

3. Labial plate with 11, 13 or 15 teeth; mandibles with obvious teeth ...... •...... ••...... •...•...... •...... 4 Labial plate with 3 teeth (may appear like 5) (Fig. 6.70); mandibles without obvious 'teeth (Fig. 6.7p) ...... Corynocera sp. [Undescribed larvae of this genus have been recorded from several lakes in New Zealand.]

4. Labial plate with 11 or 15 teeth ...••...... •...... 5 Labial plate with 13 distinct, unicolourous teeth (or 11 if the median tooth is considered trifid) (Fig. 6. 7q) ...... •..•..•...... • ...... • . . . . . '.. '. Calopsectra Kieffer sp. [All life history stages of this species, collected from the Hurunui River hot springs, await description.] 154

5. Anterior margin of labial plate strongly convex and with 11 teeth, median tooth rounded and unicolourous with no sign of notching (Fig. 6.7r) .••....•..•••.••. Tanytarsus vespertinus Hutton [This species has been recorded from lakes and rivers in lowland and upland areas. Freeman 1959, Hutton 1902] Median tooth of labial plate not unicolourous and/or not uniformly rounded ..•.••..•••..••.••.•••...•..•••..•.....•••... .• 6

6. Median tooth of labial plate with basal, lateral notches (i.e., trifid); labial plate unicolourous and strongly convex (Fig. 6.7s) .....••..•...•••.••..•. Paratanytarsus agameta (Forsyth) [This species has been recorded from shallow ponds and some lakes in the northern third of the North Island. Forsyth 1971, Kieffer 1921) Median tooth of labial plate not unicolourous and with slight lateral notching (Fig. 6.7u), which may make it appear to comprise 5 teeth in newly moulted larvae (Fig. 6.7t)i labial plate only slightly convex .••.....•• Calopsectra funebris (Freeman) [This species can be found in rivers, lakes, ponds, swamps and some oxidation ponds. Forsyth 1971, Freeman 1959, Sublette & Wirth 1980]

7. Antennae 6-segmented (Fig. 6.7e) .••••..•....•.•...•..•...••...... 8 Antennae 5-segmented ....•...••..•..•••...•.•.•.•.....•...... ••.•• 9

8. Paired median teeth of labial plate smaller than first laterals, second laterals small and on the side of third laterals, 16 teeth (Fig. 6.7d,f) (early instars may have 15 teeth, i.e., only one small median tooth) ••.•.•.•••..•••.•.•••••..•••. Paucispinigera spp. [One described species, P. approximata Freeman (Fig. 6.7d,e), and one undescribed species. The larva of the latter species (known from Lakes Gault and Matheson) has minute middle and second lateral teeth on the labial plate (Fig. 6.7f). P. approximata inhabits beech forest streams and some lakes, especially those with beech-derived organic substrates. Freeman 1959] Paired median teeth of labial plate lighter than laterals and larger than first laterals which are on the sides of second laterals (Fig. 6. 7g) •••• •...... •.•.••••. ?Microtendipes Kieffer sp. [Known from a single larva collected in Blue Lake, Tongariro.] 155

9. Labial plate concave anteriorly, the middle tooth wide and light, flanked by oblique rows of darker laterals (Fig. 6.7h). Maxillary palps prominent Cryptochironomus Kieffer sp. [Recorded from Waitomo Stream, North Island.] Labial plate not as above .•.•....••••••.•..••.•••...••.•.•.•.... 10

10. Labial plate concave anteriorly with 8 low, rounded, black teeth; paralabial plates indistinct; mandibles triangular and darkly pigmented (Fig. 6.7i) ....••••.•.. Harrisius pallidus Freeman [The larva of this species occurs inside partly decomposing wood in mountain-streams. Freeman 1959] Labial plate not as above .....••...... ••..•...••...... •.•.•.. 11

11. Labial plate with 14 teeth, the paired median and second laterals largest and even in height (Fig. 6.7j) .....•..•.... Polypedilum spp. [Ten described species and possibly several undescribed species all belonging to the subgenus Polypedilum: polypedilum pavidus (Hutton), P. longicrus Kieffer, P. opimus (Hutton), P. harrisi Freeman, P. digitulus Freeman, P. cumberi Freeman, P. ignavus (Hutton), P. canum Freeman, P. luteum Forsyth, and P. alternans Forsyth. P. ignavus may be a synonym of P. canum. Specific determination is possible only by examination of adult males. The genus is represented in a wide range of freshwater habitats: P. pavidus is common in the littoral zone of eutrophic lakes and some oxidation ponds, P. opimus and P. harrisi inhabit small streams and seepages and P. luteum probably occurs in running waters. Forsyth 1971, Freeman 1959, Hutton 1902, Kieffer 1921] Labial plate usually with an odd number of teeth (if even then greater than 14) .••••••...... ••••••••.••.•••...... •••...... •••• 12

12. Eighth abdominal segment with one or two pairs of ventral

tubules (= blood gills) (Fig. 6.6b) ...... 13 Eighth abdominal segment without ventral tubules ...•...... ••• 14

13. Two pairs of ventral tubules (Fig. 6.6b11 variable in length, usually as long as segment 8; labial plate with 15 teeth (Fig. 6.7k); two pairs of anal gills, each less than half the length of segment 8, directed posteriorly •••••.•.•.•.....• Chironomus zealandicus Hudson [This is the "thummi" type of the common, red "blood worm" and is found in the benthos of lakes, streams and eutrophic waters such as oxidation ponds. The larva of Chironomus analis Freeman appears to be morphologicallY indistinguishable from "thummi" type C. zealandicus but differs cytologically. Forsyth 1971, Freeman lS6

19S9, Hudson 1892, Hutton 1902] One pair of ventral tubules (arising distally on abdominal segment 8) with pointed ends; labial plate convex with.lS teeth (if first laterals are considered to be formed by lateral notches of the large median tooth); paralabials with finely serrated anterior edge and pointed ends (Fig. 6.71); anal gills bulbous towards apex and directed laterally .•...••••..••...... •...... Kiefferulus opalensis Forsyth [Found on wood and among roots of Juncus Spa in ponds and lakes. Forsyth 1975b]

14. Labial plate with 13 teeth, the outer lateral pair (Le., 6th

laterals) each with a slight notch, 5th latera~ small (Fig. 6.7m) ...... •...... Cladopelma curtivalva (Kieffer) [This larva, which is common in lakes, was described erroneously by Forsyth (1971) as Chironomus (Cryptochironomus) cylindricus. Forsyth 1971, Freeman 1961] Labial plate not as above, with IS or more teeth •...••.••..••..• IS

IS. Larva an obligate commensal of the freshwater mussel (Hyridella menziesi (Gray»; labial plate variable, in the 4th instar with 5-8 small similar medial teeth forming a nearly straight line flanked by a smaller separate tooth, then a large tooth beginning a descending series of7 smaller teeth. Teeth of the 3rd ins tar variable in number and disposition •••....•..••..•...••..•..••... .•...... •...... •• Xenochironomus canterburyensis (Freeman) [Probably widely distributed in lakes inhabited by the molluscan

host. Forsyth 1979, Forsyth & McCallum 1978a & b, Freeman 19S9] Larva free-living; labial plate with IS teeth ...•..••.•.•..•...• 16

16. Paralabial plates with coarsely serrated anterior margin and

recurved striations (Fig. 6.7n) ....•• ~ ...••..•.••...... ••...• Parachironomus cylindricus (Freeman) [This is the larva of Freemanis Chironomus (Cryptochironomus) cylindricus which Saether (1977) assigned tentatively to the genus Parachironomus. This, relatively uncommon, species is found in lakes. Freeman 1959] Para labial plate without serrated anterior margin: median and second lateral teeth of . similar size, first laterals smaller (Fig. 6.7k) •...•...•..••..•..•.•.•...•.•....•.••..• Chironomus sp.a [This species is the "salinarius" type of C. zealandicus Hudson and is found in many fresh-brackish-polluted water habitats. Forsyth 1971, Hudson 1892. Hutton 1902] 157

Orthocladiinae

Wise (1973) recorded 12 species in 10 genera from New Zealand but recent collecting indicates that there are many more species yet to be described. Of those in Wise's list, Camptocladius stercorarius (De Geer) is unlikely to have aquatic larvae as its immature stages occur in cow dung in Britain and parts of Europe (De Geer 1776) while the larvae of Smittia verna (Hutton) (Hutton 1902) are likely to be terrestrial like those of most other species in the genus. The larvae of only three of the other species listed by Wise are known. Those of Corynoneura donovani Forsyth (Fig. 6.7v) and Lymnophyes vestitus (Skuse) have been adequately described (Forsyth 1971, .Edward & Colless 1968) but the description of syncricotopus pluriserialis (Freeman) (Forsyth 1971) is insufficiently detailed to separate it from very similar Cricotopus species.

Because it deals with such a small proportion of the fauna and does not use generic or specific diagnostic characters to separate taxa, Forsyth's (1971) key is of limited value. Keys written for use in other countries also have little utility in. this country and it is not uncommon to find the same specimen will key to different genera in different keys. If overseas keys are used, the user should qualify his identification with a statement such as, "keys to Cricotopus Spa in Mason (1973)".

Adult chironomids, like larvae, almost always need to be mounted on slides for microscopic examination and identification. Methods for preparing and mounting specimens for examination are given by Edwards (1929), Schlee (1966), Saether (1969) and Pinder (1978).

The following key to adult male chironomids is based on the works of Pagast (1947), Freeman (1959, 1961), Beck & Beck (1966), Brundin (1966), Forsyth (1971, 1975b), Coffman (1978), Pinder (1978) and Sublette & Wirth (1980) as well as additional taxonomic investigations of my own. The key includes 26 genera and 52 species of chironomids from the subfamilies Tanypodinae, Podonominae, Diamesinae and Chironominae (including all the described species known from New Zealand). Eight further species (of the genus Parochlus (Podonominae» are included but can be identified to species only as pupae (Brundin 1966). No attempt has been made to key to genera or within the subfamilies Telmatogetoninae (marine) and the difficult Orthocladiinae. 158

head capsula abdominal segment anal setae ~.---:------.,,-'------. - ~f 'II I ( . . \\ maxI al)'pa p t---__.c~'._~_. r' _.A....C ..~ "')I.... posterior procerci ~,1 I = preanal papillae) ;t~ paralabial comb anterior prolag t Imay be absent) posterior prol6g .,,4----1...... - suspensorium a Tanypodinaa of hypophal)'nl(

anal gill b Chironomus zeafandicus

antenna I blade crenulations f seta interna :-:::"'-\>---=:::'I

labial plate paralabial para labial plate plate (striated)

9 occipital margin

Fig. 6.6 a. Tanypodinae, larva. b. Chironomus zealandicus, larva. c. Tanypodinae head capsule (ventral). d. orthocladiinae head capsule (ventral). e. Chironominae head capsule (ventral) . f. Ablabesmyia mala, labial palp. g. Parochlus sp., labial plate. h. Parochlus sp., mandible. Scale bar = 1 mID. 159

,- antennal blade

~lautarborn organ

ring organ

o

(y".".·~·t ""-

Fig. 6.7 All figures show labial and para labial plates unless otherwise indicated. a. Maoridiamesa sp. he,ad capsule, dorsal (after Brundin 1966). b. M. stouti labial plate Brundin 1966). c. Lobodiamesa campbelli labial plate (after Brundin 1966). d. Paucispinigera approximata. e. P. approximata antenna. f. Paucispinigera sp. g. 7Microtendipes (after Mason 1973). h. Cryptochironomus sp. i. Harrisius pallidus labial plate, antennae and mandibles. j. Polypedilum sp.· k. Chironomus zealandicus, C. analis and C. sp.a. 1. Kiefferulus opalensis (after Forsyth 1975). m. Cladopelma curtivalva. n. Parachironomus cylindricus. o. Corynocera SPa p. Corynocera SPa mandible. q. Calopsectra SPa r. Tanytarsus vespertinus. s. Paratanytarsus agameta (after Forsyth 1971). t. Calopsectra funebris newly moulted larva showing details of "middle tooth" (after Sublette & Wirth 1980). u. Calopsectra funebris. v. Corynoneura SPa head capsule dorsal. 160

6.3.2 Key to Adult Male Chironomidae of New Zealand

* Tribes not known from New Zealand.

Reference should be made to Fig. 6.8 for morphological termin?logy, and to figures given in published adult descriptions (referenced, where appropriate, below) to assist in the interpretation of this key.

pubesct!tJ( ey~ 8 =-___ scutltllum ____ wing base I--___postnolum

¥-___,halteltJ

plfJ-epistfJrnum

dorsa-central setae_--,U:...a1, o ilfIUlliiGijll'- tibial comb scutellar selae ___lJl{j;. ....:.;tr--- last ngment tibial spur of tarsus postnOtum __---'~~ E median 9roove __---l~'LJ

cosla fl, ..

.::tr-__ claw

gonoco,;le lobe squamal fringe G An f.Cu M I-m m-cu

anallergile bands

K

l+---''M-/¥-__ appendage 1 appendage 1a -"d---'~~

Fig. 6.8 A. Male antenna. B. Part of the head of a chironomid showing its pubescent eye. C. Lateral and D. dorsal views of the thorax of an adult chironornid. E. Tibial combs and spurs of a chironominae adult. F. Foot of a chironomid showing large pulvilli. G. Wing of a tanypodinid showing terminology applied to veins. Dorsal views of male hypopygia of H. Tanypodinae, I. orthocladiinae, J. Chironornini and K. Tanytarsini. (Figures from Pinder 1978) 161

1. Postnotum without a median groove or keel; antennal filament consisting of 4-6 segments (marine) ...... , ....•.. Telmatogetoninae (not keyed) Postnotum usually with a median groove or keel (except Tribe Clunionini, an exclusively marine tribe of Orthocladiinae) ; antennal filament consisting of at least 10 segments ..•..•.....• 2

2. X - vein M-Cu present...... 3 X - vein M-Cu absent ...•...•....•..•...... •...•••.•....•...•.. 5

3. R2+3 present ...... •...... •...•.•••.•.•.....• ,...... 4

R2+3 absent Podonominae ...... • • . • • • • • • • • • •• 6

4. R2+ 3 simple Diamesinae ...... •...... 7 R2+ 3 forked Tanypodinae •...... •..... 13

5. Gonostylus directed rigidly backwards .•••..• Chironominae ..... 20 Gonostylus directed inwards Orthocladiinae •.. (not keyed)

6. Gonystylus with two lobes .....•.. 'Podonomini ...•...... •...... 22 Gonostylus simple ...... • Boreochlini*

7. f-Cu distal to M-Cu X-vein ....•...... •...•.•... 8 f-Cu proximal to M-Cu X-vein, OR M-Cu X-vein runs into f-Cu .... 10

8. Antepronotal lobes prolonged and strongly curved outwards; mesothorax with expanse of short, suberect hairs arising from large pits •..•...•..•..•.... Lobodiamesini [One species known: Lobodiamesa campbelli Pagast 1947. Pagast 1947, p. 443, Fig. 1; Brundin 1966, p.417, 603-608. ] Antepronotum normal; mesothorax with single, double or triple row of bristles ...•..•..•...... ••.••••..•.•..••....••...... 9

9. Eye fuscous (= 'hairy') ..•. Heptagyini (in part) Maoridiamesa ... 32 Eye bare ...... ••.•• Prodiamesini*

:10. Gonostylus narrow, and attached me one third the way up from the distal end of gonocoxite ..•..... Protanypiini* Gonostylus attached at distal end of gonocoxite ...... 11

11. Gonostylus short and robust ...... Boreoheptagyini* Gonostylus longer and narrower ....•.•...... ••...•...... •. 12

12. M-Cu X-vein runs into f-Cu ..•.•..... Heptagyini (in part) * f-Cu proximal to M-Cu Diamesini*

13. 4th tarsal segment cordiform (= heart-shaped) ..• Coelotanypodini* 4th tarsal segment cylindrical...... 14 162

14. f-Cu sessile •..... (i.e. proximal or opposite M-Cu X-vein) ..•• 15 f-Cu stalked ...•... (i.e. distal to M-Cu X-vein) ..•.•...•...... 19

15. costa not, or only slightly, produced beyond the end of R4+5 , .• . . . , •.••... Pentaneurini in part (except Natarsia) ..•...... 35 Costa produced beyond the end of R4+5 by 2X length of r-m .•...• 16

16. Wings with macrotrichia (larger hair-like structures on wing surface, cf. microtrichia) at distal end only; no rows of proximal spines on tibial spurs ..•...... Anatopyniini* wings with dense macrotrichia; tibial spurs with rows of proximal spines .•..•.••.•..•...•..•.•.••....•....••••.•..•.••• 17

17. Eyes irridescent •...•..•..•• Macropelopiini in part {Psectrotanypus & Apsectrotanypus)*? ..•. 36 Eyes not irridescent ••....••.••...•...... ••..•••...... •••.... 17

18. Mesonotal tubercle (= hump-like plate lying in middle of

mesonotum) present i claws pointed ...••••.•..•••••.•••••.•••.•. ••.•. Macropelopiini in part (Macropelopia) ....•••.••...•..... 36 Mesonotal tubercles absent; claws much split at apices .•.•..•...••••..• Pentaneurini in part (Natarsia)*

19. stalk of f-Cu less than one third as long as CU2i eyes irridescenti mesonotal tubercle present ••...... • Tanypodini*

stalk of f-Cu more than half as long as CU2i eyes black; mesonotal tubercle absent .•.•••. Macropelopiini in part (procladius & Psilotanypus)*

20. Wing with macrotrichia on membrane and X-vein r-m.parallel to

and practically continuous with R4+5 i squama bare •.••..•.••.•• • • . • . • • • . • . • . . • . . • . . • . . . .• Tanytars ini .••...•.•.....•••••.•... 45 [N.B. Adult of Corynocera sp. not known] Wing usually without macrotrichia on membrane, when present then r-m transverse; squama usually fringed ...... ••.•..•...•.• 21

21. Anterior tibia with well-formed black spur with enlarged base . • • • • . . . • . . . . . •• Pseudochironomini [One described species: Riethia zeylandica Freeman 1959 Freeman 1959, p. 423, Figs 3a and 3a'] Anterior tibia without spurs or with only a small spur or scale •.•.•...••...••..•.. Chironomini ••.•••..••..•••••.••••••. 48

22. 4th tarsal segment with wide membranous sole extending from distal margin ...... •...•..•• Podonomus ...... •..•.... 25 4th tarsal segment without sole ...... •.•...••••...•••..• 23 163

23. Ganastylus ventrally with small labe; segment 15 .of antenna langer than 14; eyes densely fuscaus •...... •.•..•...... • ...... • . . . . . • • • . •• • • . .. Padachlus " ..' ...... •..... 27 Ganastylus nat as abave; segment 15 .of antenna sharter than 14; eyes bare •••••••••••••..•.••••••• . .• . . . • . • • . • • . • • • • • • • •• 24

24. Last palpal segment strangly swallen; .outer spur of hind tibia nearly as lang as inner; wings very strongly reduced ..•.•.•..• ...... • ...... • . . . .. Zelandochlus [One described species: Z. latipalpis Brundin 1966 Brundin 1966, p. 105, Fig. 35, p. 107, Figs 36-39.] Last palpal segment not swallen; .outer spur much shorter than inner or lacking; wings nat reduced ...••• Parochlus 30

25. Apical labe .of ganastylus long and slender, situated in a distinctly ventral positian relative to the subapical labe; basal part of gonastylus strangly swallen ••...•.•.••••••••••. ...•...... •.•••••...... Podanomus parochloides Brundin 1966 [Brundin 1966, p. 201, Fig. 173] Apical lobe .of ganastylus shart, situated on level .of subapical labe •....••.••••••.••.•.•.•...•...•..•..••...... •...•...•...•. 26

26. Apical labe widened terminally ..• Podonomus waikukupae Brundin 1966 [Brundin 1966, p. 201, Fig. 172] Apical labe not widened terminally Podonomus pygmaeus Brundin 1966 [Brundin 1966, p. 201, Fig. 174]

27. Inner labe .of ganastylus with lang and slender terminal taath •. 28 Inner lobe .of ganostylus with shart and broad terminal taath ••• 29

28. Antennal segment 14 1.5X langer than 13, segments 10-13 cylindrical .•..••••••••...... •.•.• Podochlus grandis Brundin 1966 [Brundin 1966, p. 249, Fig. 289] Antennal segment 14 samewhat shorter than 13, segments 10-13 battle-shaped ••.••..•.•.....•...•.. Podochlus stouti Brundin 1966 [Brundin 1966, p. 249, Fig. 290]

29. Inner lobe .of ganastylus parallel-sided

...... "''''' Q" ".0" (I Q" ... It ...... " .. Podochlus cockaynei Brundin 1966 [Brundin 1966, p. 249, Fig. 291] Inner lobe .of ganastylus with marked swelling subterminally •.. Podochlus knoxi Brundin 1966 [Brundin 1966, p. 249, Fig. 292] 164

30. Cell Rl of wing of normal width (i.e., maximum width of cell Rl approximately equal to length of r-m X-vein) .•.••.•.•••.••..•. Parochlus conjungens Brundin 1966 [Brundin 1966, p.129, Fig. 52] Cell Rl narrow (i.e., maximum width of cell Rl approximately half length of r-m X-vein) •....•...•..•..•.•••.••.•.....•.•••. 31

31. Mesonotum well arched; wings and legs of normal type; subapical lobe of gonostylus rather slender, 2-3X longer than broad ..••. •...•..•...•.. 8 spp. of Parochlus recognisable only as pupae. [P. aotearoae Brundin 1966, P. spinosus Brundin 1966, P. maorii Brundin 1966, P. ohakunensis (Freeman 1959) (Freeman 1959, p. 403, Fig. lc},P. carinatus Brundin 1966, P. pauperatus Brundin 1966, P. novaezelandiae Brundin 1966, P. longicornis Brundin 1966] Mesonotum only slightly arched; wings very narrow; legs conspicuously long; subapical lobes of gonostylus very large, only a little longer than broad ... Parochlus glacialis Brundin 1966 [Brundin 1966, p. 157, Fig. 113]

32. Legs with hair of normal length; male antennae 11-14 segments .. 33 Legs with very short hair; body light coloured; male antennae 7-segmented; gonostylus very large, much longer than gonocoxite ...... •...... •.. Maoridiamesa insularis Brundin 1966 [Brundin 1966, p. 397, Fig. 551]

33. Gonostylus compact, little more than 2X longer than broad ••..• Maoridiamesa harrisi Brundin 1966 [Brundin 1966, p.395, Fig. 549] Gonostylus slender, 3.5- 4x longer than broad •••••••.•..••••.• 34

34. Male antennae l4-segmented with normal plume ••.•••.•••.•.••. ,. Maoridiamesa stouti Brundin 1966 [Brundin 1966, p. 397, Fig. 550] Male antennae II-segmented, reduced plume •.••••.•••...••.••.•• Maoridiamesa glacialis Brundin 1966 [Brundin 1966, p. 414]

35. Tibiae with 3 or 4 well-defined black rings; prescutellar area well-defined, more or less circular with acrostichal bristles diverging around it ...•.•..••.••.• Ablabesmyia mala (Hutton 1902) [Freeman 1959, p. 400; Sublette & Wirth 1980, p.303-304, p. 338, Fig. 4] 165

Tibiae without black rings; pres~utellar area not well-defined and with acrostichal bristles running across it •••.... Pentaneura [One described species P. harrisi Freeman 1959 although more are known to exist. Freeman 1959, p. 400]

36. Dark femoral markings when present, confined to an apical or subapical ring; pulvilli absent ••....••.•.•.•••..•••.•..... 37 Femora with central as well as one or two subapical dark rings; small pulvilli present ...... ••...... •...•.•...•...... 44

37. Wings with a discrete dark spot on centre of stem of posterior

fork (= vein Cu proximal to f-Cu); a dark i wings heavily marbled; postnotum bare .... Gressittius antarcticus (Hudson 1892) [Freeman 1959, p. 402 & Plate II, Fig. a; Sublette & Wirth 1980, p.302 & p.335, .1] Wings without a dark spot on centre stem of posterior fork; postnotum with a group of 6-10 hairs .•••••••••••••••••.•.•.... 38

38. Tarsal segments lacking dark tips .•••••..•••••.•••••.•••...••• Macropelopia (Freeman 1959) [Freeman 1959, p. 405, Fig. lb] Tarsal segments with dark tips ...•.....•••.....•..•••..•...••• 39

39. Wings with obvious pale spots near wing margin in cells Ml and M3 ..•••.•.••..••.••.•..•....•.•.....••••..•..•••...•.•••.•••.. 40

Wings without pale spots in cells Ml and M3 , often uniformly clouded or more or less clear, or with discrete dark spots at apices of veins ••••••••••.•••••.•..•.•••••..•••..•••••••..•••• 42

40 Dark markings on anterior margin of abdominal segments (in

dorsal v~ew) ••••••••.•••.•. Macropelopia apicincta (Freeman 1959) [Freeman 1959, p. 403 & Plate II, Fig. b] Dark markings either basally or centrally placed on dorsal surface of abdominal segments •.•.....•.•.....•.•••..••••••.. . 41

41. M3+4 with an elongate dark cloud along most of its length; anal cell with a single large rectangular dark patch .....•...... •.•...... Macropelopia debilis (Hutton 1902) [Freeman 1959, p. 404 & Plate II, Fig. d] M3+4 dark at tip; anal cell with two separate dark patches •••• .•...... ••..•..•••.•....•.•• Macropelopia languidus (Hutton 1902) [Freeman 1959, p. 403 & Plate II, Fig. c] 42. wing with five distinct dark spots ...••.•....•••••...•..•..••...... Macropelopia quinquepunctata (Freeman 1959) 166

[Freeman 1959, p. 405, Fig. la] Wing without distinct dark spots at apices of veins M3+4 and Cu 1 ....•...••.•..••.••.•. •..••••••.•••.•.•...•.. •.•.•.••...• 43

43. Wing apex with broad cloud containing a dark spot near middle of cell R5; abdominal pattern formed of a row of three spots on each segment ...••...... •• Macropelopia umbrosa (Freeman 1959) [Freeman 1959, p. 406 & Plate II, Fig. e) Wing apex less distinctly clouded, no darker spot in cell R5' abdominal pattern formed of a basal or sub-basal band on each segment, usually absent from segments 1 and 2 •.•••..••.•.••••• • ...... • ...... • Macropelopia apicinella (Freeman 1959) [Freeman 1959, p. 406]

44. Mesonotum pruinose (with dull surface) beb"een the stripes; anal cell with a pale area within the dark at the tip; wing pattern darker; femur dark at base with two subapical rings .....•...... •.• ?Apsectrotanypus quadricincta (Freeman 1959) [Freeman 1959, p. 407 & Plate II, Fig. f] Mesonotum with whitish pruinosity allover; anal cell dark at apex and without included pale area, wing pattern paler; femur pale at base and with only one subapical ring ...... •.•.•.•...• ?Apsectrotanypus cana (Freeman 1959) [Freeman 1959, p. 408]

45. Combs of posterior tibiae well separated, at least one bearing a longish spur .•.••••.••..••.•.....•...••..•.....••.•.....•..• 46 Combs of posterior tibiae usually contiguous, with or without spurs; if clearly separated, spurs are absent ...•..••...... •.. 48

46. Anal point with 16 irregularly arranged spinulae (small spines or thorn-like projections), appendage la present ...... •...... Tanytarsus albanyensis Forsyth 1971 [It is possible that this species should be placed in the genus Calopsectra but I have not seen any specimens. Forsyth 1971, p. 138, Fig. 14d] Anal point with fewer spinulae arranged in a single/double row, appendage la absent...... •..•.. 47

47. Wing length 2.0 - 2.5 mm; anal point with a single row of 3 to (perhaps) 6 spinulae ••••••.•• Calopsectra funebris (Freeman 1959) [Freeman 1959, p. 436, Fig. 6b; Sublette & wirth 1980, p. 374, Fig. 40c] 167

Wing length about 1.6 rom; anal point with only one or two spinulae ..•.....•••••..•....••.•.•.... Calopsectra undescribed sp. [This undescribed species is known from the Hurunui River hot

springs area (= Micropsectra in Stark, Fordyce & Winterbourn 1976) and morphologically is very similar to C. funebris.** C. funebris however, is a much larger species and has dark brown thoracic markings whereas the undescribed species is smaller and paler (light yellow-brown). Associated larvae differ in the form of the labial plate (see larval key) . Adults of both species resemble the Australian Tanytarsus fuscithorax Skuse 1889 and T. inextentus Skuse 1889.]

Footnote ** Selected measurements from the undescribed Calopsectra and C. funebris. Number examined given in brackets.

Palp ratios 0.038:0.084:0.099:0.176 rom 0.039:0.110:0.117:0.214 mm

AR* 0.91 - 0.92 (2) 0.67 - 1.14 (5) (3)_Freeman 1959 WL* (rom) 1. 59 - 1.63 (3) 2.00 - 2.18 gave WL = 2.5 m:ll VR* 1.13 - 1.15 (2) 1.06 - 1.19 (3)

Macrotrichia on wing veins R with 16 - 18 (2) 22 - 27 (3) Rl with 12 - 15 (2) 21 - 25 (3) R4+5 with 15 - 16 (2) 21 - 26 (3) Leg ratios (LR*) Fore 1.88 - 1.90 (4) 1. 71 - 1.89 (3) Mid 0.49 - 0.53 (8) 0.53 - 0.54 (3) Hind 0.61 0.64 (4) 0.61 - 0.63 (3)

Total body length (rom) 2.10 3.28

* AR antennal ratio = the ratio of the length of the last two segments (Tanypodinae) or last one (other subfamilies) of the antennal flagellum to the short basal segments of the flagellum taken together. WL wing length. Cont'd 168

VR = vena rum ratio ~ a ratio obtained by dividing the length of Cu, measured from its base (at the arculus) up to the cubital fork (f-Cu) by the length of M, measured from its base (at the arculus) up to r-m. LR leg ratio = the ratio of the basitarsus (i.e., the basal or proximal segment of the tarsus) to the tibia. Unless otherwise stated, this ratio applies only to the front legs.

48. Combs on hind tibiae fused and armed with one or two spurs .•.. Paratanytarsus agameta (Forsyth 1971) [Female only known. Forsyth 1971, pp.137-l41] Combs separated and unarmed •.• Tanytarsus vespertinus Freeman 1959 [Freeman 1959, p. 436, Fig. 6a]

49. Posterior tibiae with two spurs (i.e., each comb with a spur) .. SO One spur on small outer comb, large inner comb unarmed .....••• 58

50. Pronotum much reduced; mesonotum projects as a cone over the head .•...... ••.•..•..•.••••••..••..•....•...... •••..•. 51 Pronotum reaching up to front of mesothorax, sometimes collar­ like, sometimes closely applied to mesonotum 52

51. Wing membrane thickly clothed with macrotrichia ..••...•...•... Harrisius pallidus Freeman 1959 [Freeman 1959, p. 427, Fig. 4a] Wing membrane without macrotrichia ..•...... •...... ••...... Ophryophorus ramiferus Freeman 1959 [Freeman 1959, p. 427, Fig. 4b]

52. Appendage 2 in male genitalia extremely broad and bulbous apicallY (i.e. tennis racquet-shaped) ••••••...••.•••....••••...... Kiefferulus opalensis Forsyth 1975 [Forsyth 1975b, p. 216, Fig. la] Appendage 2 more slender, not or scarcely enlarged distally ... 53

53, Appendage 2 reaching well beyond tip of gonocoxite .f

Appendage 2 very much reduced or absent e ••• G .... O ...... "' ...... , ...... 5b

54. Anal point very broad at distal end, not tapering to a point; appendage 1 short, broadly ovate and pubescent •...

...... •... ~ ...... Xenochironomus canterburyensis (Freeman 1959) [Freeman 1959, p. 425; Forsyth 1971, p. 126, Fig. 7c] Anal point narrower, tapering to a point; appendage 1 well developed, chitinised and bare except for a few setae basally

.... • ...... "' ...... 0 ..... 00 Chironomus spp...... •....•...... •• 55 169

55. Anal point of male narrow, not much wider basally than near tip ••.••••••••...•.•••...... Chironomus zealandicus Hudson 1892 [Freeman 1959, p. 423, Fig. 3b] Anal point stout, much wider basally than near tip ••••....•... Chironomus analis Freeman 1959 [Freeman 1959, p. 423, Fig. 3c]

56. Appendage 1 rod-like and bearing a few setae Parachironomus (Freeman 1959) [2 apical setae. Freeman 1959, p. 423, . 3d] Appendage 1 reduced ...... •.....•....•••••••••••••.••...... •.. 57

57. Appendage 1 reduced and bearing 3 apical setae; gonostylus long and curved ••.••••••••.•..••. Cladopelma curtivalva (Kieffer 1917) [Freeman 1961, p. 697, Fig. 21f] Appendage 1 short, broad and pubescent; gonostylus short and broad ...•.•.••••••••••.•..•...••.....••••••• ?Cryptochironomus sp. [Adult of N.Z. larva keying to this genus not known]

58. Squama bare ••.•.•..••.•.•...•. Microtendipes/Paralauterborniella? [Adult of N.Z. larva keying to this generic complex not known] Squama • • . • • • • • . . • . . . . . • . • • • • • . • • • • • • • • • • • • • • • • • • ...... 59

59. Pulvilli easily visible and divided longitudinally; anterior tibial scale ( tibial comb with 'teeth' usually with small spur; 8th abdominal segment of male constricted basally ..•.•.•...... •..•.•.•....••...... polypedilum spp •.••.•...•... 60 Pulvilli only visible on slide mounts, not divided; anterior tibiae without either scale or spur; 8th tergite not constricted basally ••.•..•.••••••.•...... Paucispinigera spp. • •...... 68

60. Wings with dark markings and clouds ..•••.••••..•....••.••.•.•. 61 Wings unmarked .•••..•••••.••.•.•...•••••••.•••••..•.•...•••••. 63

61. R4+5 strongly curved; veins CUI and M with macrotrichia •••..•...... •...... • ...... • Polypedilum opimus (Hutton 1902) [Freeman 1959, Plate Ill] R4+5 practically straight; veins CUI and M bare •••.•••...•.•.. 62

62 . Wing length 1. 3 - 1.5 mm; wing markings more de f ini te and including a dark spot basal to the X-vein .•.•.....•.•...... •.•...... •...... polypedilum longicrus Kieffer 1921 [Freeman 1959, p. 431, Fig. 5b, Plate Ilk] 170

Wing length 3.5 4.0 rom; markings ill-defined, dark spot basal to X-vein absent .•..•.••.••...• Podypedilum pavidus {Hutton 1902} [Freeman 1959, p. 431, Fig. 5a, Plate IIj]

63. Abdomen dark with pale markings or pale with dark markings; costal cell rounded at tip ..•.•..••.••.....•...... •.••.... ,... 64 Abdomen dark brown or black without pale markings; costal cell pointed ..•.•..•..•...••...•..••....•.....••.••.•...•.•..••.... 67

64. Cylindrical finger-like process between bases of gonocoxites of male ...... •.....••....•.••..•• polypedilum digitulus Freeman 1959 [Freeman 1959, p. 431, Fig. 5e] Finger-like process absent •.••••....••..•.•..•..••.••.....•... 65

65. Abdominal segments dark with pale markings along anterior margins .•.•.••...•.••••....•.•.. Polypedilum harrisi Freeman 1959 [Freeman 1959, p. 431, Fig. 5d] Abdomen pale {yellow} or pale with dark markings (yellow/brown) 66

66. Wing length 1.4 rom; anterior tibial scale with short spur ..•.. . •• •...... • •• . . . . • . •• . .•. . •. . •. Polypedilum luteum Forsyth 1971 [Forsyth 1971, p. 135, Figs. 12a & 12b] Wing length 1.8 - 2.0 rom; anterior tibial scale without obvious spur •..•...... ••.•.....•. polypedilum alternans Forsyth 1971 [Forsyth 1971, p. 138, Figs. 14a & 14c]

67. Abdomen blackish and without pruinose bands ....•..•.•..•...... Polypedilum cumberi Freeman 1959 [Freeman 1959, p. 431, Fig. Sf] Abdomen dark brown or blackish with pruinose bands at apices of segments •....•...... •...... •.• Polypedilum canum Freeman 1959 [Po ignavus (Hutton) may be a synonym of P. canum, Freeman 1959, p. 431, Fig. 5g]

68. Appendage 1 blunt-ended •• Paucispinigera approximata Freeman 1959 [Freeman 1959, p. 427, Figs 4c & 4d] Appendage 1 pointed apically ..•...• Paucispinigera undescribed sp. [See Fig. 6.9. Adult known only from male pupa collected from L. Gault, South Island.] 171

Fig. 6.9 Male hypopygium of undescribed Paucispinigera sp. (one side only shown).

(Scale bar 0.05 nun)

6.3.3 Description of the Adult Male of Eukiefferiella sp. (Chironomidae: Orthocladiinae)

Head dark brown, palps lighter, similar to legs. Thorax dark, scuta and pre-epipisternum shining dark brown-black. Legs uniform mid-brown. Abdomen brown and hairy, somewhat darker on distal margins of segments.

Eyes bare and without dorsal extension. Antenna with 13 flagellomeres, its apex distinctively shaped (Fig. 6.l0B). AR = 1.00- 1.06, mean 1.02 (n = 3).

Wing (Fig. 6.l0A) with very fine microtrichia, WL = 1.97 - 2.13 rom, mean 2.05 rom (5), VR = 1.37 - 1.40, mean 1.39 (4). Squama fringed with approximately 8 setae. Costa not produced, R4+5 reaches wing margin proximal to wing tip. nearly straight. R with macrotrichia (approximately 6 - 8) • R2+3 feint and lying close to R4+5 but without clear ending in costa. M reaches wing margin at wing tip. CU2 almost straight but curved close to wing margin and not distinctly reaching wing margin. R4+5 ending above the end of CUI' Anal vein short and curved down near tip, not quite reaching f-Cu. F-Cu well beyond r-m. Anal lobe well developed.

Thorax. Scutellum with a single transverse row of 8 setae.

Scuta1 stripes fused, dorsocentral (= dorso-lateral setae of some authors) setae (approximately 8) erect and uniserial. Mesonotum without visible acrostichal setae.

Legs. LR = 0.50 - 0.53, mean 0.52 (5). Distal end of hind tibiae expanded, outer spine of hind tibiae less than half as long as 172 inner, 5th tarsal segment slightly dorsoventrally flattened, small pulvilli present and empodium (e.g. Fig. 6.8F) well developed. Legs with distinct hairs (up to 0.20 rom long).

Abdomen covered with hairs, segments 7 and 8 constricted basally.

Hypopygium (Fig. 6.lGC). Anal point absent. Gonocoxite lobe free and well developed, gonostylus expanded medially with a terminal triangular point and strong subterminal tooth, prominent seta distal to base of medial expansion. A second prominent seta arises from end of gonostylus just distal to terminal triangular point.

Fig. 6.10 Eukiefferiella sp. male A. Wing B. Terminal of antenna (large hairs arise from the distinct pits - only one is shown). C. Hypopygium (large hairs, which arise from the distinct pits, and fine hairs found on gonocoxite lobe are not shown). Scale bars are 0.05 rom (B & C) and 0.5 rom (A). 173

Specimen localities

South Island. BR Lake Rotoiti, Nelson, 21 November 1978, B.V. Timms 200; MC - Lake Grasmere, 1976- 1980 900; Kenwyn Avenue, Christchurch, 7 November 1979 10; Avon River near junction with Okeover Stream, Christchurch, 22 November 1979, V.M. Stout 10.

Diagnosis

This fits the diagnoses of Eukiefferiella Thienemann 1926 as given by Coe (1950) and Sublette & Wirth (1980) which are broader than the diagnosis given by Brundin (1966) who stated that Eukiefferiella never had pulvilli and that R4+5 ended proximal to CU1' As now recognised, this genus is an artificial group of heterogeneous species (Sublette & Wirth 1980). The combination of a normally developed antepronotum, weakly bowed CU2' R2+3 usually lying near R4+5 or completely fused with it, the end of R4+5 over or proximal to the end of CU1' outer spur of hind tibia weakly developed or absent, weak or absent dorsomedial (= acrostichal) setae, and genitalia usually without an anal point and with a glossiform (= tongue-shaped) medial lobe to the gonocoxite, differentiates most species of the genus.

The described here keys to Psectrocladius Kieffer in Brundin (1956) and Pinder (1978) unless the pulvilli are considered 'small and difficult to see'. They are small but distinct. Further, this species would be referred to the subgenus Mesopsectrocladius Laville since it possesses dorsoventrally flattened fifth tarsal segments and lacks an anal point (Langton 1980). The above keys separate Psectrocladius and Eukiefferiella on the presence or absence, respectively, of distinct pulvilli (a rather unsatisfactory couplet). However, this New Zealand species differs from the Psectrocladius diagnosis (Brundin 1956) in that R2+3 does not end about midway between the ends of and and f-Cu is markedly distal to r-m, and is, therefore, more properly placed in the genus Eukiefferiella.

6,3.4 Chironomid Larvae and from Lake Grasmere

The purpose of this section is to document the 'types' of chironomid larvae (and some pupae) collected from Lake Grasmere and to outline briefly the bases for their recognition. Figure 6.11 shows diagnostic features of the chironomids that presented problems. 174

J N : :. !. , . ,1; . I)' .-;~':. )l. .' H I

M K L~

Fig. 6.11 Morphological features of some Chironomidae larvae and pupae from Lake Grasmere. Caudal swim fins of pupal (A) Macropelopia languidus and (B) M. umbrosa; paralabial combs of pupal (C) Gressittius antarcticus-type and (D) ? M. umbrosa; pupal respiratory trumpets of (E) Syncricotopus pluriserialis and (F) Cricotopus zealandicus; labial plates of (G) S. pluriserialis and (H) Cricotopus Sp.i (I) mandible and (J) antennae of Cricotopus sp.; labial plates and antennae of (K) Orthocladiinae A and (L) Orthocladiinae B ?Rheocricotopus sp.; (M) labial plate of Orthocladiinae C. [Scale bars 0.05 rom except

(1) Tanypodinae

Three kinds of tanypodine larvae were recognised in quantitative samples from Lake Grasmere: Ablabesmyia mala, Pentaneura sp. and Macropelopia/Gressittius (two species).

A. mala did not present problems of specific identification. Larvae were recorded from Lake Grasmere (Appendix 3) and adults were collected in light-trap and hand-net samples from the lake shore (Appendices 5.1 and 5.2).

Only one species of pentaneura (P. harrisi Freeman) has been recorded from New Zealand (Freeman 1959, Forsyth 1971). Adults of this species are known from the Cass area (checklist in Burrows 1977) but, in the absence of a positive adult association, I hesitate to identify the two larval specimens collected further than to genus. The genus Pentaneura (sensu Freeman 1961) is now considered to be a composite group of genera (Martin 1974) and there is some doubt as to the correct generic placement of Australasian Pentaneura (Dr J. Martin pers. comm.). I know of the existence of adults of an undescribed New Zealand species of 'Pentaneura' and I have also seen larvae and pupae (different from the Grasmere types and P. harrisi) from the Reef ton area of the South Island (see Cowie 1980).

New Zealand larvae keying to the tribe Macropelopiini also present problems of identification. Ten species have been described as adults (Hudson 1892, Hutton 1902, Freeman 1959, Sublette & wirth 1980) but the larva of only one (Gressittius antarcticus (Hudson» has been described (Forsyth 1971), and since larvae of most of the other species are unrecognised, it is not known whether this description adequately distinguishes this species from congenerics.

Three species of adult Tanypodinae in the tribe Macropelopiinae were recorded from Lake Grasmere and surrounding areas (Appendix 5.2) but only two pupal types were collected from the lake. The most common pupa was that of Gressittius antarcticus (established from reared association) (see Forsyth 1971, pp.116-ll7, Fig. 2; Sublette & wirth 1980, pp. 303 & 335-336, Figs 1 & 2). The respiratory trumpet of this species, with its wide and convoluted trachea, is very distinctive (see Sublette & wirth 1980, p. 335, Fig. IE). Adults of G. antarcticus were also the most common Macropelopiini collected (Appendices 5.1 & 5.2). 176

Adults of two further species of Macropelopia (M. umbrosa and M. languidus) were found in light-trap samples from the southern end of Lake Grasmere (Appendix 5.2). However, the absence of M. languidus in hand-net collections from the lake shore (Appendix 5.1) and the low numbers caught in light-traps (Appendix 5.2) suggests that larvae and pupae were the least likely to be present in the lake. They may have been present in nearby Lake Pearson or in streams.

Fig. 6.11 shows the caudal swim fins of the pupae of M. languidus (reared from the Avon River, Christchurch) and M. umbrosa (reared from Lake Grasmere) . The caudal region of the pupa of M. languidus (Fig. 6.llA) is similar to that of G. antarcticus (cf. Sublette & wirth 1980, Fig. 2E) but is much smaller (Table 6.1) and the respiratory trumpet is of the M. apicinella-type (cf. Forsyth 1971, p. 118, Fig. 3b) with broad, apically-directed spinules on the external surface. Pupae of the M. languidus type have not been collected from Lake Grasmere.

Table 6.1 Maximum lengths and widths of respiratory trumpets and caudal swim fins of pupae of Gressittius antarcticus, Macropelopia languidus and M. umbrosa.

G. antarcticus M. languidus M. umbrosa

Respiratory trumpet length x width (mm) 0.75 x 0.22 0.48 x 0.22 0.36 x 0.17

Caudal swim fin length x width (both 1.31 x 1.37 0.89 x 0.80 0.69 x 0.62 fins together) (rom)

The caudal region of the pupa of M. umbrosa (Fig. 6.llB) differs from those of G. antarcticus and M. lanquidus in that the swim fins taper to a more acute point, but the respiratory trumpet is similar in shape to the apicinella-type although the apically directed spinules are minute. Pupae of M. umbrosa were collected from Lake Grasmere.

Two separate types of Macropelopiini larvae were distinguished from material collected during the quantitative sampling program. They differed in the number of teeth in the paralabial comb (Fig. 6 .1lC & D) • The paralabial comb of G. antarcticus (Forsyth 1971, p. 117, Fig. 2b) is not unlike that depicted in Fig. 6.llC. Although pupae of 177

G. antarcticus and M. umbrosa have been collected and reared to adults from Lake Grasmere l positive associations with larvae have not been made. As these two species were common in light-trap and hand-net collected samples of adults (Appendices 5.1 & 5.2), it is possible that the two larval types recognised are M. umbrosa (Fig. 6.11D) and G. antarcticus (Fig. 6.11C). However, in the absence of reared associations with adult material, the two larval morphs were not separated in the analyses.

(2) Podonominae

Adults of Parochlus, belonging to the Araucanus group (see Brundin 1966), were found in light-trap and hand-net collections from the shores of Lake Grasmere (Appendices 5.1 & 5.2). Eight species in New Zealand belong in this group but are recognised clearly only in the pupal stage (Brundin 1966). No work has been done in an attempt to separate adults, but those collected from Lake Grasmere were most likely of one species since no characters could be found to separate them. Larvae and pupae of Parochlus were not found in quantitative samples from the macrophyte zones of Lake Grasmere but pupae of P. spinosus Brundin 1966 were collected from the stony shore at the southern end of the lake. This species is in the Araucanus group (in the spinosus subgroup) so it is likely that the adults collected were of this species also.

(3) Orthocladiinae

Orthocladiinae presented the greatest problems of generic and specific identification and many of the difficulties could not be overcome.

The Syncricotopus/Cricotopus complex Positively identified adults of Syncricotopus pluriserialis and Cricotopus zealandicus were recorded in hand-net and light-trap collections from the lake shore (Appendices 5.1 & 5.2) and pupae of these species were found in the lake. The respiratory trumpets of S. pluriserialis and C. zealandicus differ in size and ornamentation (Fig. 6.llE,F). S. pluriserialis larvae (from ponds near the main entrance of the James Hight Library at the University of Canterbury) (labial plate, Fig. 6.l1G) have been reared to adults to confirm specific identity and several Cricotopus-type larvae from Lake Grasmere 178

(labial plate, Fig. 6.11H; mandible, Fig. 6.111) were reared and shown to be C. zealandicus. However, there was much variation in the form of the labial plate (for example, in the width and shape of the middle tooth) in larvae from Lake Grasmere keying to the genus Cricotopus and it is uncertain whether this variation is inter- or intraspecific. There was also variation in the relative lengths of the antennal segments (Fig. 6.11J) but this was probably inter-instar variability,

Larvae of the Cricotopus-type were the most common chironomids on macrophytes in Lake Grasmere (Appendix 3) but only a single individual of S. pluriserialis (a pupa) was recorded on macrophytes. The tube­ dwelling larvae of S. pluriserialis were present, however, amongst filamentous algae and diatoms on the upper surface of rocks on the stony shore at the southern end of the lake, although still outnumbered by larvae of Cricotopus sp. (at least during summer). It was somewhat anomalous, therefore, that adults of S. pluriserialis were much more common than those of C. zealandicus in light-trap and hand-net collections (Appendices 5.1 & 5.2). But this may be explained by the fact that S. pluriserialis is a multivoltine species (Forsyth 1971) whereas C. zealandicus is probably univoltine. The peak of C. zealandicus emergence was probably not recorded by the relatively infrequent hand-net sampling of adults (Appendix 5.1) and this species did not seem to be attracted appreciably to light (Appendix 5.2).

Hirvenoja (1973), in a revision of the genus Cricotopus and its closest relatives, synonymised the Australasian Syncricotopus Brundin 1956 with the genus Paratrichocladius Santos Abreu 1918. Whereas the adults of New Zealand Syncricotopus agree with this diagnosis of Paratrichocladius, the larvae do not. S. pluriserialis larvae from New Zealand have a seta interna on the inner margin of the mandible, a feature that is absent in Paratrichocladius. Because of this, it is my belief that Syncricotopus and paratrichocladius are not synonyms and, therefore, I consider that the New Zealand (and Australian) species should be retained in Syncricotopus.

Orthocladiinae A

The larva of Orthocladiinae A (Fig. 6.11K) keyed to Psectrocladius in Mason (1973) but did not key satisfactorily using the more regionalised keys of Bryce & Hobart (1972) and Oliver, McClymont & Roussel (1978). It is possible that this larva is the immature stage of the adult Eukiefferiella sp. described previously (see p. 171). 179

Orthocladiinae B

The larva of Orthocladiinae B (labial plate and antenna Fig. 6.11L) keyed to Rheocricotopus in Oliver, McClymont & Roussel (1978); Trichocladius in Mason (1973); and Rheocricotopus (Trichocladius) in Bryce & Hobart (1972). I did not manage to associate this larval type positively with an adult, and could not find any adults in hand-net or light-trap collections that possessed the required features (see the generic diagnosis of Rheocricotopus in Brundin (1956). Larvae of this type have been recorded from the benthos of Lake Grasmere (and several other South Island lakes) by Timms (1980, in prep) .

Orthocladiinae C

I tentatively attribute this unusual larva (labial plate Fig. 6.11M) to the subfamily Orthocladiinae, but it could represent a new subfamily. Superficially, the dark pigmentation of the head is reminiscent of Maoridiamesa (although it lacks the pronounced black occipital margin) but the labial plate is not of the Maoridiamesa­ type (see the larval key, p. 149). The flattened, forked 'hairs' comprising the beard on the paralabial plates are unlike those of any known chironomid larvae. Overseas keys to larvae were of no value in identifying this species.

The unidentified 'black orthoclad' (Appendix 5.1) with a patterned wing may be the adult of Orthocladiinae C (presumptive association), and, as with the larva, overseas keys were of no use for its identification. For example, it keyed to Eukiefferiella in Pinder (1978) but then did not agree with the generic diagnosis given by Coe (1950), Brundin (1956) or Sublette & Wirth (1980). The pupa of this adult (positive association) differed from the normal orthoclad type in lacking stout bristles on the caudal fins (or anywhere on the abdomen) and was more like a diamesid in this respect.

Description of the adult male of the 'black orthoclad'

WL = 1. 6 - 1. 7 mm, AR = O. 4, LR = O. 5 - O. 6 Antenna 13-segmented, plume sparse; eyes hairy. Dorsocentrals (5) uniserial, erect and arising from distinct pits. Wing without macrotrichia but with very obscure black markings across middle, in anal cell and nearer wing tip; squama bare; Costa not produced and ending on a level with CUI' R2+3 close to R4+5 and ending 180 in Costa near the end of R4+5, f-Cu well beyond r-m, CU2 gently curved (but not straight), M straight and ending at wing-tip. Legs. All tarsi of normal shape, the 5th perhaps slightly flattened dorsoventrally, pulvilli absent and small empodium present. Hypopygium. Anal point distinct, sharp and bare, gonocoxite without a medial lobe but with a slight ridge with prominent setation, peg on gonostylus appears continuous with gonostylus (i.e., seems to be made of the same material and not articulated).

(4) Chironominae

Larvae of two species of Chironominae, Chironomus zealandicus (Tribe Chironomini) and Tanytarsus vespertinus (Tribe Tanytarsini) were recorded in quantitative samples. All life history stages of both species were recognised easily. 181

CHAPTER VII

GENERAL DISCUSSION

Few studies have been made on the macrophyte-associated macro- invertebrates in New Zealand lakes, although the characteristic of this habitat have been documented in general surveys (see for example, stout 1975b, Winterbourn & Lewis 1975). Taxonomic limitations in past years make comparisons of species lists difficult (Table 7.1), however, allowing for this, it appears that the number of invertebrate taxa recorded from the survey of Lake Grasmere is the recorded from any New Zealand lake. About 113 macroinvertebrate species were collected from the lake and its immediate environs, of which about 87 were associated with macrophytes. Of these, 49 were insects, with of Chironomidae (17) and Trichoptera (12) best represented. Thirty invertebrate species were new records for the Cass district (c.f. checklist in Burrows 1977). In New Zealand lakes, species of crustacea and Trichoptera are represented at least as well as in many overseas lakes but Ephemeroptera, Odonata, Hemiptera, Coleoptera, Acarina, and Mollusca are poorly represented (Table 7.1). Comparisons of chironomid species richness are hampered by taxonomic difficulties, but in the Nearctic and Palearctic, where midge faunas are relatively well known, the littoral zone of a mesotrophic (= moderately productive) lake may be inhabited by some 50 chironomid species (see saether 1979) . In contrast, the total New Zealand freshwater chironomid fauna comprises approximately 70 described species (Wise 1973, Stark in press) with perhaps another 10 -15 species undescribed (mostly Orthocladiinae) (Stark in press and unpublished records). Of these, perhaps 40% may live in standing waters. The 17 recorded from Lake Grasmere ly the largest number recorded from a New Zealand lake) may be close to the maximum species richness for such a habitat.

Stout (1975b) noted that the freshwater fauna of Canterbury (and New Zealand as a whole) comprised fewer than would be found in comparable geographic regions in the northern hemisphere, and that several groups (e.g., freshwater Anostraca and Conchostraca, Polyphemidae, Leptodora, Diaptomus, Asellus, and insects living on the water surface such as the hemipterans Nepa, and Ranatra and Gyrinidae) are absent or poorly represented. Similarly, Winterbourn (1980) noted that ponds and Table 1.1 Taxonomic comparisons of macrophyte-associated macroinvertebrate f~unas of Lake Grasmere, New Zealand and other lakes in New Zealand ~nd

overseas. Also given is the taxonomic composition of the benthos of Lake Grasmere (Timms pers. corom.). w • not recorded. - E not present.

.- Species composition L. Grasmere, L. Waitaki, L. Aviemore, Three Dubs Tarn, Hodson's Tarn, Great Grebe L., L. Erken, Goczalkowice 12 lakes in by taxa N.Z. N.Z. N.Z. England England Sweden Sweden Reservoir. Snowdonia, benthos (Greig 1973) (Greig 1973) (Macan 1949) (Macan 1963) (Berg & Petersen (Nyman Poland Wales 195G) 1971) (Kuflikowski (Liddle eC al. 1974) I 1979)

Coelenterata 1 - - - - - 1 1 2 - Platyhelminthes 1 1 1 - 2 1 3 1 - 2 Ectoprocta 1 - - - - - 2 1 1 - Nematoda 2+ - 1+ - - - - 1+ 1 1 Annelida 7 4. 7 7 5 4+ 2+ 13 21 3 Crustacea 16 - 2 8 9 1 1 3 I 9 1 Ephemeroptera 1 - 1 1 3 4 1 7 9 4. Odonata 3 1 1 1 5 7 9 2 5 6 P1ecoptera 3 - - - 1 1 - - - 1 Hemiptera 2 - - - 10 8 - 2 3 G Coleoptera 4. - 2+ 2+ 7 8 G 4 12 :2 All Diptera 23 4. 5+ 4+ 3+ 1+ 2+ 4+ 29 * Chironomidae 17 4. 3+ 2+ 1+ .. R * 22 .. Neuroptera ------2 - - Trichoptera 12 4 5 G G 4 9 23 12 5 Lepidoptera 1 1 1 - - 1 - 1 3 1 Acarina 5 1 1 1 16 13 12 .. 10 1 Mollusca 5 3 6 4 2 1 - 15 15 3

Total all species 87+ 19 32+ 33+ 69+ 43+ 48+ 80+ 132 36+ Total insects 49+ 10 15+ 14+ 35+ 33+ 25+ 45+ 85 26+ I I 183

lakes in New Zealand are geologically young habitats that support an impoverished fauna. Thus, in some northern European eutrophic (= nutrient-rich) lakes it is not uncommon to find 155 invertebrate species on wave-exposed stony shores (Ehrenberg 1957), 300 species in the macrophyte zone (Muller-Liebenau 1956),50 species in the sublittoral, and up to 20 species in the profundal (Jonasson 1978) (cf. Lake Grasmere, Table 7.1).

As noted for other lakes (by, for example, Muller-Liebenau 1956, Gerking 1957, Soszka 1975a, Wetzel 1975, Jonasson 1978, Higler 1980, and Morgan 1980), the invertebrate fauna of the macrophyte zone of Lake Grasmere was more diverse than that of the benthos (Table 7.1) and all species found in the bottom mud (Dr B.V. Timms pers. comm.) were present also on aquatic plants. The species richness of the benthos of seven Rotorua lakes (North Island) (a mean of 12 taxa per lake including eight insects, but Annelida were not differentiated) studied by Forsyth and McColl (1974) and Forsyth (1975a, 1978) was similar to that recorded for Lake Grasmere (Dr B.V. Timms pers. comm. and Table 7.1). Timms (1980) also studied the benthos of the Nelson lakes (South Island) where the community composition was similar to that of Lake Grasmere benthos (except that more species of Chironomidae were present).

In addition to differences in the taxonomic representation of lacustrine invertebrates between New Zealand and the northern hemisphere, there are differences in the groups that are numerically dominant. Mollusca (55.4%), primarily P. antipodarum, Coelenterata (16.1%) and Crustacea (15.5%) were best represented in the macrophyte-associated invertebrate communities in Lake Grasmere. Insecta (4.2%), including Trichoptera (2.1%) and Chironomidae (2.0%), Acarina (3.8%), and Annelida (4.4%) comprised most of the remainder. Dominance of macrophyte- associated invertebrate communities by Mollusca (principally P. antipodarum) has been noted also for other lakes in New Zealand by, for example, Greig (1973) (Lakes Aviemore and Waitaki) and Stout (1975b) (Canterbury lakes). In contrast, invertebrate communities associated with macrophytes in northern hemisphere lakes usually are dominated by Diptera (especially Chironomidae), Annelida (especially Oligochaeta), and sometimes Ephemeroptera (Table 7.2, Gak et ale 1972, Morgan 1980). Only a few overseas lakes (e.g., Lakes Chad and Lere in Africa and some warm temperate lakes in the Soviet Union) are dominated by Mollusca (Morgan 1980). 184

Table 7.2 Dominant invertebrate groups in macrophyte zones of some northern hemisphere lakes.

Lake Erie, U.S.A. Annelida 39.4%, Chironomidae 23.9% (Krecker 1939) Goczalkowice Res., Poland Chironomidae 64-82%, Oligochaeta 9-12% (Kuflikowski 1974) Lake Velence, Hungary Chironomidae 80-87%, Ephemeroptera 4-12% (Andrikovics 1975) Mikolajske Lake, Poland Diptera 50%, Oligochaeta 35% (Soszka 1975a) Dubh Lochan, Scotland Ephemeroptera 40-45%, Oligochaeta 33-36%, (Minto 1977) Chironomidae 13-18%

Differences in invertebrate abundance were observed on different macrophyte species in Lake Grasmere. Of the plants examined (Elodea canadensis, Myriophyllum propinquum, Isoetes alpinus and Ranunculus

fluitans) only the last named could be considered a rel~tively poor habitat, with invertebrate densities usually less than half those on the other plants (in terms of numbers per g dry weight of plant or numbers per m2 of lake bottom). I. alpinus supported a relatively abundant fauna including many animals (e.g., Ostracoda and some Chironomidae) that were not actually on the plants but were living in the silty substrate that collects in the root zone and around stem bases. It was impossible to separate these animals from those living on the plant itself. On all plants the principal features of their associated animal communities were"the high contributions made by molluscs (59-86% by numbers) and the greater contribution of insects to the faunas on native (M. propinquum 9.5%, I. alpinus 8.4%) compared with adventive (E. canadensis 2.4%, R. fluitans 2.9%) macrophytes. However, all of the species that comprised greater than 0.1% (by numbers) of total invertebrates collected were present on all four plants. The apparent macrophyte-substrate specificity of less common taxa was probably a function of their rarity rather than an expression of their niche specificity.

Invertebrate population densities on different macrophytes have been found to bear some relationship to morphological features of the plants (Krecker 1939, Andrews & Hasler 1943, Entz 1947, Rosine 1955, Gerking 1957, Greig 1973, cattaneo & Kalff 1980), or they may be related to water chemistry (Frost 1942), chemical composition of plants, or difference in plant periphyton (Harrod 1964, Wolnomiejski pers. comm. in 185

Soszka 1975a, Cattaneo & Kalff 1980). In Lake Grasmere, the macrophytes with the most finely divided leaves (M. propinquum and E. canadensis), perhaps by providing increased protection and surface area for periphyton growth and invertebrate grazing, facilitated the presence of a more diverse and abundant fauna than did R. fluitans. Cattaneo & Kalff (1980) found, by experiment, that plants with finely dissected leaves (e.g., Myriophyllum) accumulated significantly more periphyton biomass than plants with relatively simple leaves (e.g., Potamogeton) and that this was reflected in the densities of their associated invertebrate populations. However, many authors (e.g., Godward 1937, Flint 1950, Macan 1963, Pieczynska & Spodnieska 1963, Macan & Kitching 1972, 1976, Soszka 1975b, Wetzel 1975, Hodgkiss & Tai 1976, Bowker & Denny 1978) have noted marked similarity in periphyton community composition on different species of macrophyte, artificial plants and stony substrates, and often a marked similarity in species composition of the associated invertebrate fauna. Therefore, differences in composition of periphyton-browsing invertebrate communities on different macrophyte species and between macrophytes arid stony/rocky substrates are probably due primarily to factors other than food availability (e.g., wave action, oxygen saturation, water depth). In Lake Grasmere, most of the macrophyte-associated invertebrates were found also on stony or rocky substrates. Notable exceptions included Chlorohydra viridissima and Nymphula nitens (almost exclusively on macrophytes) and Procordulia grayi (in silty substrates). However, a number of species were virtually restricted to the stony shore zone (although some of these were collected rarely from plants) (e.g., Deleatidium sp., Stenoperla prasina, Austroperla cyrene, Zelandobius furcillatus, polyplectropus puerilis, psilochorema nemorale). These mayfly, stonefly and caddisfly species are more commonly found in running waters but are on exposed shores in lakes winterbourn & Lewis 1975).

The introduced macrophyte E. canadensis, a native of North America, has been in'New Zealand for over 100 years (Mason 1975, Hughes 1976) . It is not known when it was introduced into Lake Grasmere although it was present in nearby Lake Sarah in 1934-1935 (Flint 1938). E. canadensis formed a dense monoculture in mesotrophic Lake Grasmere and excluded all other macrophyte species at depths ranging from about 0.5- 1.2 m to 7 m. The lower limit is within the photic zone and compares with a maximum Secchi disc visibility of 8.2 m recorded by Stout (1977). Brown (1975) stated that the upper depth limit of E. canadensis is determined by the wave exposure pattern of a lake and 186

the lower limit is related to light penetration or, perhaps, temperature. Coffey (1975) contended that when E. canadensis competes with native submerged macrophytes it coexists with, rather than displaces, them in

oligotrophic (= nutrient poor) and eutrophic waters, whereas in mesotrophic waters it displaces the natives completely within its depth range (as in Lake Grasmere). Further spread in this lake is unlikely unless land-use changes lead to eutrophication or environmental conditions change dramatically (e.g., long periods of calm conditions allowing E. canadensis to colonise shallow water).

It is difficult to determine the precise effects that the displacement of native macrophytes by E. canadensis has had in Lake Grasmere since no macrophyte or invertebrate surveys had been undertaken prior to its introduction. Coffey (1975), from an intensive study of macrophyte communities in the waikato lakes (North Island), suggested that the dominant elements of the native, submerged macrophyte floras were Myriophyllum and Potamogeton spp. and charophyte algae in oligotrophic waters, with Potamogeton becoming increasingly dominant in more mesotrophic conditions. Brown (1975) noted that the widespread occurrence of comparable species throughout New Zealand suggests that a similar native species association might be a common feature of lakes not markedly affected by adventives. Therefore, charophytes and Potamogeton may have been the original deep-water vegetation in Lake Grasmere, with mixed native species of Myriophyllum, Potamogeton and Isoetes in shallow water (see stout 1975b and references therein) . Native communities of tall-growing Potamogeton and Myriophyllum species have fewer stems per unit area of lake bottom than adventives like E. canadensis (Brown 1975), and potamogeton in particular, may exhibit pronounced die-back during winter (Stout 1975a). Since almost all invertebrate species collected from E. canadensis in Lake Grasmere were present on the other macrophytes studied and on other substrates (e.g., sand, stones or rocks), it is probable that invertebrate species richness has not been reduced by displacement of the native plants. Furthermore, since invertebrate densities per m2 of lake bottom were higher on E. canadensis than on the native plants, the total invertebrate population of the lake may now be greater.

The invertebrates associated with aquatic macrophytes in Lake Grasmere exhibit a range of feeding strategies. It is often impractical or impossible to separate species into rigid trophic levels (Morgan 1980) since most aquatic invertebrate species are generalist (polyphagous) rather than specialist feeders (Cummins 1973, Merrit & 187

Cummins 1978, Cummins & Klug 1979, Mackay & Wiggins 1979), and food habits may be subject to considerable intraspecific variation such as age-specific or habitat-specific differences (Cummins 1973, Anderson & Cummins 1979, present study). Despite this, most can be assigned to "most probable feeding categories" and this enables invertebrate community composition of macrophytic and benthic habitats of some New Zealand lakes to be considered in terms of functional groups (modified from Cummins 1973) as in Table 7.3. or higher taxa, were assigned to functional feeding groups on the basis of their predominant food type and feeding mechanisms. For Hudsonema amabilis was termed an herbivorous browser since, although it was strictly omnivorous, most of its food was diatoms, macrophyte tissue and filamentous algae. On the other hand, Nymphula nitens and cephalotes fed primarily on macrophyte tissue and were classed as shredders, even though other food types were recorded in their faeces. The common snail Potamopyrgus antipodarum, was considered an herbivorous browser when living on plants or stones but a detritivorous browser when on the benthic mud in the profunda 1 zone.

Table 7.3 Feeding types (% by numbers) in relation to habitat in some New Zealand lakes.

L. Grasmere L. Waikaremoana1 L. Rotoiti2 L. Rotoroa2 Macrophyte zone Benthos

Shredders 0.2 0.1 0.2 Browsers Detritivores 17.4 94.6 76.5 76.5 Herbivores 77 .5 2.3 10.4 4.7 Predators 4.6 0.3 5.0 9.8 Filterers 0.3 2.8 0.1 8.8

(Data from Main (1976)1, Timms (1980)2 and present study)

macrophytes seem to be little used as food in running waters (Westlake 1975, Anderson & Sedell 1979) or lakes (Entz 1947, Rosine 1955, Soszka 1975b, Jonasson 1978, Cattaneo & Kalff 1980) and this is attributed to their high C/N ratios, large quantities of relatively indigestible cellulose and lignin, and low digestibility of their (Boyd 1970). In the macrophyte zone of Lake Grasmere, shredders of aquatic macrophytes (viz., N. nitens and T. cephalotes) 188 comprised only 0.2% of the invertebrate community (Table 7.3). They were poorly represented also in the benthos (Dr B.V. Timms pers. comm.) and in the benthos of other New Zealand lakes (Table 7.3, Forsyth 1975). Herbivorous browsers dominated the macrophyte zone of Lake Grasmere (Table 7.3) and included gastropod molluscs, hydroptilid caddisflies,

H. amabilis J Eucyclops serrulatus J and most orthocladine chironomids. Detritivorous browsers (mostly oligochaetes and oribatid mites) were next in abundance (Table 7.3). Since most of the browsing invertebrates recorded on macrophytes in Lake Grasmere were found in other habitats as well (e.g., on stones and/or mud), it seems that macrophytes are used mainly as a substrate from which periphyton is grazed. Predators (the swimming mites, tanypodine chironomids, some Trichopterae.g., Oecetis spp., Antiporus strigosulus, and Odonata) comprised 4.6% of the macrophyte-associated invertebrate community in Lake Grasmere, and filterers (bivalve molluscs) only 0.3% (Table 7.3). In the profundal benthos of New Zealand lakes, communities are dominated by detritivorous browsers (mainly Oligochaeta), and filterers tend to be better represented than in macrophytic habitats (Table 7.3, Forsyth 1978) . Although quantitative data on the composition of the benthos of Lake Grasmere is not immediately available, semiquantitative information (Dr B.V. Timms pers. comm.) indicates that detritivorous browsers (Oligochaeta, P. antipodarum, and Chironomidae) dominate the fauna, as in the benthos of other lakes (Table 7.3, Jonasson 1978).

Proportions of different feeding groups (e.g., detritivores, shredders, predators, filterers) may vary at different depths within a lake (Jonasson 1978, Table 7.3). In macrophyte-rich localities the accumulated remains of plants are decomposed through a detritus food chain (Wetzel & Allen 1972) since utilisers of fresh macrophyte tissue are rare (Cummins & Klug 1979, Cattaneo & Kalff 1980). On the rocky shore of a lake where wave action prevents the growth of aquatic macrophytes but diatoms and other algae may grow on the rocks, grazing food chains predominate (see Jonasson 1978). In the profundal benthos, invertebrate communities are dominated by detritivorous deposit feeders (e.g., most oligochaetes and some chironomids) and filter-feeders (bivalve molluscs) that feed on accumulated fine particulate organic matter (Jonasson 1978).

In Lake Grasmere, detritus, living macrophytes, filamentous algae, diatoms and animal tissues probably represent an increasing nutritional gradient (Anderson & Cummins 1979), and the food habits of most invertebrates studied were characterised by ingestion of food 189 of increasing quality with increasing age. For example, the or changes in the food habits of H. amabilis with age were the reduction in detritus intake and increased consumption of macrophyte tissue and animal prey items. T. cephalotes and N. nitens also showed relatively less ingestion of detritus and increased ingestion of macrophyte tissue in later instars. N. nitens was observed to feed mainly on the young shoot tips which would have had rather limited exposure to colonisation by diatoms and microorganisms. However, growing tips are more nutritious than older growth since they have greater nitrogen and protein content and less indigestible lignin (Feeny 1970, 1976). The other shredder studied (T. cephalotes), preferred older leaves and did not eat fresh growth also Babington 1967). Like many other shredders, it may derive its nutritional requirements from ingested microorganisms and macrophyte tissues that have already been partially hydrolysed by periphytic communities present on decomposing macrophytes (Cummins & Klug 1979).

A notable feature of the food habits of three of the species studied was their predation upon Mollusca (viz., P. antipodarum) The feeding of X. zealandica on P. antipodarum is the only instance of dragonfly predation upon a mollusc that I have found, and the incidence of predation on P. antipodarum by the two trichopterans (H. amabilis and T. cephalotes) also appears to be the highest recorded. Slack (1936) noted that two of 12 Molanna angustata (Trichoptera: Molannidae) examined had fed upon Lymnaea sp. in a scottish lake, and Winterbourn (1978b) found a radula.of P. antipodarum in the gut of one (of 17) psilochorema bidens (Trichoptera: Rhyacophilidae) in a New Zealand stream. Experimental studies have shown that living P. antipodarum are preyed upon actively by H. amabilis (Wilson 1980) and Neolimnia spp. (Diptera: Sciomyzidae) (Barnes 1979) both of which enter the shell and can prise aside the operculum. This seems to be a very effective feeding strategy and, considering the marked dominance of molluscs in many New Zealand freshwater habitats (Winterbourn & Lewis 1975, present study), makes available an abundant food resource.

In the macrophyte zones of Lake Grasmere, the fauna at all sites was dominated by browsers, mainly those feeding on periphyton growing on plant surfaces, but a number of species' populations had their primary representation in different areas of the lake. For example, although P. hendersoni was the most common hydroptilid caddisfly on plants in most parts of the lake, P. tillyardi was the most common 190

species on E. canadensis in the eastern sampling area, and o. albiceps, although sometimes recorded on plants, was most abundant on algal covered stony substrates. On plants, spatial separation may occur also at the microhabitat level. For instance, some species (like P. antipodarum) may browse predominantly on macrophyte surfaces close to the lake bottom, whereas other macrophyte-associated invertebrates may be distributed more evenly over the plants (e.g., the hydroptilids), or are present mainly on shoot tips (e.g., N. nitens).

Fig. 7.1 is a generalised food web for invertebrates in the macrophyte zone of Lake Grasmere summarising the interrelationships between functional feeding groups of invertebrates. The shredder and filterer based pathways are relatively unimportant compared with the detritus and periphyton browsing fqod chains in the macrophyte zone. Most macrophyte tissue is degraded via microbial and fungal decay and enters the detritus food chain (Wetzel & Allen 1972).

LIGHT

MACROPHYTES O",~I::TOPLANK~~7VTON

DETRITUS 4--4----:------r~--+-- Allochthonous inputs

SEDIMENTS

BROWSERS SHREDDERS Detritivores Herbivores Hud.om, ... A.I!iI<\!hlJi.· 'rrJplectldulJ c .. ptwlote.1l

Hympl'ltd .. nl tBns

rri_JoIICWlOthru. I'IOWII lflld$Dl"l.fl'at ..-.bl,u_" Pot4.l'l'Opvn,\UJ .ant.Jp0d4rue" ot't:.ho<:lbdiinae

PREDATORS

Fig. 7.1 Food web for invertebrates in the macrophyte zones of Lake Grasmere showing the importance of browsing invertebrates and the periphyton and detritus-based food chains. Prominent examples of invertebrates in each of the functional feeding groups are shown. (* denotes invertebrates that belong in more than one functional group.) 191

Highly seasonal life-history patterns are characteristic of most holarctic invertebrate species (Hynes 1970 and references therein). In contrast, in New Zealand (and other temperate regions of the southern hemisphere) many freshwater insects have weakly synchronised life­ histories characterised by poorly-defined cohort growth and long flight periods Babington 1967, Norrie 1969, Winterbourn 1974, Towns 1976, McFarlane 1977, Crumpton 1979, Deacon 1979, Cowie 1980, present study). Only a few examples of freshwater insects with well-synchronised life­ histories have been recorded. in New Zealand (e.g., Rakiura vernale, Michaelis 1973; P. tillyardi, present study). The life-histories of only a few non-insectan freshwater invertebrates have been studied in New Zealand. Winterbourn (1970b) investigated the population dynamics of P. antipodarum, Towns .(1976) the life-history of the amphipod Paracalliope fluviatilis, and there have been a number of studies of planktonic crustaceans (e.g., Green 1974, Chapman, Green & Jolly 1975, Burns 1979). Complete life-histories are known for only a few mites (viz., Eylais waikawae, piona uncata exigua, Hydrachna maramauensis, stout 1953a & b). Most freshwater mites probably have annual with adults present from late spring until autumn (stout 1976) •

The life-histories of most insects studied in Lake Grasmere (P. tillyardi, H. amabilis, T. cephalotes, P. aureola, O. unicolor, N. nitens) were univoltine, fitting Hynes' (1970) slow seasonal cycle category. Larvae recruited in late summer/autumn and emerged in spring and summer and some (e.g., H. amabilis, T. cephalotes and N. nitens especially) exhibited little or no growth during winter. On the other hand, O. albiceps and P. hendersoni may belong more properly in Hynes' (1970) non-seasonal category since individuals of all stages may be present at all times of year. Hopkins (1976) found that O. albiceps was bivoltine (with pupae in October and June), but it may be that this species and P. hendersoni have multiple overlapping flexible generations depending upon prevailing environmental conditions (especially temperature which is related to latitude). X. zealandica, since it can have a two or three-year life-history and a summer diapause in the F-2 to F instar larvae (Deacon 1979, present study) also fits Hynes' (1970) non- seasonal category.

Most work on the life-histories of New Zealand freshwater invertebrates has concerned stream insects (see Chapter V) and several explanations have been put forward to explain their observed aseasonality or poorly-synchronised life cycles. Devonport and 192

Winterbourn (1976) considered that New Zealand's relatively mild climate could be responsible, but Towns (1976) suggested that the long period of isolation of the fauna (together with severe long-term climatic variations) and the lack of pronounced autumnal leaf fall were key factors (cf. Anderson & Cummins 1979). Cowie (1980) considered various arguments in some detail and concluded with respect to stream insects that "severe and unpredictable climatic conditions would favour selection for opportunism, and consequent spreading of the risks of emergence through poorly synchronised life histories".

In Lake Grasmere, seasonality of life-histories is probably not keyed to food availability since there are no obvious periods of food shortage or pulses of abundance during the year. Rather, the relatively buffered lacustrine environment (with respect to climatic influences, especially temperature) which would not exert any selective pressure for marked seasonality, could be involved in the relatively aseasonal life-history patterns seen in most New Zealand lacustrine insects. The resultant long adult flight periods spread the risks of emergence so that there is a reduced likelihood of bad weather affecting emergence success, adult abundance, survival and mating, and consequent recruitment to the next generation. Nevertheless, marked changes in population densities may occur between years (e.g., N. nitens and X. zealandica, present study), changes that may well be related to the influences of climate on the adult stage of the life-history.

Finally I would like to emphasise the importance of sound taxonomic knowledge to ecological investigations. As Watt (1979) stated "The increasingly recognised need for accurate identifications in applied research thus often had to be met by the applied research worker first undertaking a taxonomic revision of whatever group of pests or beneficial organisms he was studying, before he could make further progress." In the present work, deficiencies in taxonomic knowledge of many species (notably Chironomidae) were a limitation until late in the study when taxonomic investigations were undertaken and 'difficult' species were identified. Much important information on, for example, habitat preferences may be overlooked if species cannot be distinguished. Further necessary taxonomic work remains to be done, especially on chironomids in the subfamily Orthocladiinae and will enable more refined ecological work to be done. 193

ACKNOWLEDGMENTS

I thank my supervisors Dr V.M. Stout and Dr M.J. winterbourn for their invaluable assistance at all during the preparation of this thesis, and for tolerating my frequent time-consuming digressions from the main

I am to the following people for assistance in identifying Drs M.J. Winterbourn, V.M. Stout (Acarina, Crustacea), G.W. Ramsay (Acarina), M.A. Chapman (Ostracoda), and Messrs P.M. Johns (Tipulidae) and A.G. McFarlane (Trichoptera). Dr J.P. Leader (Hydroptilidae) and Drs J. Martin, D.J. Forsyth and M.A. Chapman, and Mr J.A.T. Boubee (Chironomidae) kindly provided material and useful comments on taxonomic matters.

The assistance of my brother Andrew, Brent Cowie, Richard Rowe, Marianne Moore, John Preece, Ken Deacon, Bruce Warburton and Trevor Baker on sampling , and the support and encouragement of Yvonne Hall and Eric Hamilton, especially during the later stages of thesis production, is much appreciated.

Special thanks is due to the Zoology Department photographic technicians, Mrs Joan Buckley and Mr T.P. Williams for their excellent work and to Mr A. Gall for his expert technical assistance during construction of the cylinder-sampler.

Professor W. Stephenson (University of Queensland) kindly performed computer analyses on data from the main quantitative sampling program, and Dr C.L. McLay and Mr K.W. Duncan were of considerable help with statistical problems.

Finally I wish to thank Pauline Taylor for the excellent job she has done of typing the manuscript.

This work was supported by grants from the North Canterbury Acclimatisation Society and the Drummond Fund of the Church of New Zealand, whose financial assistance is gratefully acknowledged.

195

REFERENCES

ANDERSON, N.H. & CUMMINS, K.W. 1979. Influences of diet on the life histories of aquatic insects. Journal of the Fisheries Research Board of Canada 36: 335-342.

ANDERSON, N.H. & SEDELL, J.R. 1979. Detritus processing by macro- invertebrates in stream ecosystems. Annual Review of Entomology 24: 351-377.

ANDREWS, J.D. & HASLER, A.D. 1943. Fluctuations in the animal populations of the littoral zone in Lake Mendota. Transactions of the Wisconsin Academy of Science, Arts and Letters 35: 175- 185.

ANDRIKOVICS, S. 1975. Macrofaunal biomass in the submerged vegetation stands of Lake Velence. Pp. 247-254 in Salanki, J. & Ponyi, J.E. (Eds) Limnology of Shallow Waters. Akademiai Kiado, Budapest. Symposia Bio1ogica Hungarica 15: 304p.

ARMSTRONG, J.S. 1935.' Notes on the biology of Lake Taupo. Transactions of the Royal Society of New Zealand 65: 88-94.

BABINGTON, A.M. 1967. The systematics and ecology of three species of Caddisfly larvae in Lake Rotorua. Unpublished M.Sc. thesis, Victoria University, Wellington.

BARBER, H.G. 1962. Freshwater diatoms from Cass, South Island, New Zealand. Journal of the Quekett Microscopical Club 29: 17-20.

BARNES, J.K. 1979. Biology of the New Zealand genus Neo1imnia (Diptera: Sciomyzidae) . New Zealand Journal of Zoology 6: 561-576.

BARTON, D.R. & HYNES, H.B.N. 1978. Seasonal study of the fauna of bedrock substrates in the wave zones of Lakes Huron and Erie. Canadian Journal of Zoology 56: 48-54.

BAY, E.C. 1972. An observatory built in a pond provides a good view of aquatic animals and plants. The Amateur Scientist. Scientific American 227: 114-118.

BEALL, G. 1939. Methods of estimating the population of insects in a field. Biometrika 30: 422-439.

BECK, W.M. Jr & BECK, E.C. 1966. Chironomidae (Diptera) of Florida. 1. Pentaneurini (Tanypodinae). Bulletin of the Florida State Museum (Biological Science) 10: 305-379. 196

BERG, K. 1941. Contributions to the biology of the aquatic moth Acentropus niveus (Oliv). Vidensk. Medd. fra Dansk naturh. Foren. Bd. 105: 59-139.

BERG, K. & PETERSEN, I.C. 1956. Studies on the humic acid Lake Gribs¢. Folia Limnologica Scandinavica 8: 273p.

BOWKER, D.W. & DENNY, P. 1978. The periphyton communities of Nyumba ya Mungu reservoir, Tanzania. Biological Journal of the Linnean Society 10: 49-65.

BOWNIK, L.J. 1970. The periphyton of the submerged macrophytes of Mikolajskie Lake. Ekologia Polska AlB: 503-520.

BOYD, C.E. 1970. Amino acid, protein, and caloric content of vascular aquatic macrophytes. Ecology 51: 902-906.

BRADSHAW, J.D. 1977. The geology of the pre-Quaternary rocks around Casso Pp. 53-65 in C.J. Burrows (Ed.) Casso Department of Botany, University of Canterbury. 418p.

BRINKHURST, R.O. 1965. The biology of Tubificidae with special reference to pollution. Biological Problems in water Pollution, Third Seminar 1962. U.S. Department of Health, Education and Welfare. Pp. 57-64.

BROWN, J.M.A. 1975. Ecology of macrophytes. Pp. 244-262 in V.H. Jolly & J.M.A. Brown (Eds) New Zealand Lakes. Auckland University Press/Oxford University Press. 388p.

BRUNDIN, L. 1956. Zur Systematik der Orthocladiinae (Dipt. Chironomidae). Fishery Board of Sweden. Institute of Freshwater Research, Drottningholm 37: 1-185.

BRUNDIN, L. 1966. Transantarctic Relationships and their Significance as Evidenced by Chironomid Midges. Almquist & Wiksell, Stockholm. 472p.

BRYCE, D. & HOBART, A. 1972. The biology and identification of the larvae of the Chironomidae (Diptera). Entomologists Gazette 23: 175-217.

BURNS, C.W. 1979. Population dynamics and production of Boeckella dilatata (Copepoda: Calanoida) in Lake Hayes, New Zealand. Archiv fur Hydrobiologie (Monographische Beitrage) Supplementband 54: 409-465.

BURROWS, C.J. (Ed.) 1977. Cass, History and Science in the Cass District, Canterbury, New Zealand. Department of Botany, University of Canterbury, Christchurch, New Zealand. 418p. 197

CATTANEO, A. & KALFF, J. 1980. The relative contribution of aquatic macrophytes and their epiphytes to the production of macrophyte beds. Limnology and Oceanography 25: 280-289.

CHAPMAN, M.A. 1963. A review of the freshwater ostracods of New Zealand. Hydrobiologia 22: 1-40.

CHAPMAN, M.A. 1976. Ostracoda. Pp. 74-100 in Chapman, M.A. & Lewis, M.H. (Eds), An introduction to the freshwater Crustacea of New Zealand. Collins, Auckland. 261p.

CHAPMAN, M.A., GREEN, J.D. & JOLLY, V.H. 1975. Zooplankton. Pp. 209- 230 in V.H. Jolly & J.M.A. Brown (Eds), New Zealand Lakes. Auckland University Press/Oxford University Press. 388p.

CHAPMAN, M.A. & LEWIS, M~H. 1976. An Introduction to the Freshwater Crustacea of New Zealand. Collins, Auckland. 261p.

CHUTTER, F.M. 1961. Certain aspects of the morphology and ecology of the nymphs of several species of Pseudagrion Selys (Odonata). Archiv fur Hydrobiologie 57: 430-463.

CLIFFORD, H.T. & STEPHENSON, W. 1975. An Introduction to Numerical Classification. Academic Press, New York. 229p.

COE, R.L. 1950. Family Chironomidae. Pp. 121-206 in: Handbooks for the Identification of British Insects 9(2). 216p. Royal Entomological Society of London.

COFFEY, B.T. 1975. Macrophyte distribution in the Waikato Lakes. Pp. 263-270 in V.H. Jolly & J.M.A. Brown (Eds), New Zealand Lakes. Auckland University Press/Oxford University Press. 388p.

COFFMAN, W.P. 1978. Chironomidae. Pp. 345-376 in Merritt, R.W. & Cummins, K.W. (Eds), An Introduction to the Aquatic Insects of North America. Kendall-Hunt, Dubuque, Iowa. 441p.

COFFMAN, W.P., CUMMINS, K.W. & WUYCHECK, J.C. 1971. Energy flow in a woodland stream ecosystem. I. Tissue support trophic structure of the autumnal community. Archiv fur Hydrobiologie 68: 232-276.

CORBET, P.S. 1957. The life history of the emperor dragonfly, Anax imperator Leach (Odonata: Aeshnidae). Journal of Animal Ecology 26: 1-69.

COWIE, B. 1980. Community dynamics of the benthic fauna in a West Coast stream ecosystem. Unpublished Ph.D. thesis, Department of Zoology, University of Canterbury, Christchurch, New Zealand. 200p. 198

COWLEY, D.R. 1978. Studies on the larvae of New Zealand Trichoptera. New Zealand Journal of Zoology 5: 639-750.

CROSBY, T.K. 1973. Dyar's rule predated by Brooks' rule. New Zealand Entomologist 5: 175-176.

CROSBY, T.K. 1975. Food of the New Zealand trichopterans Hydrobiosis parumbripennis McFarlane and Hydropsyche colonica McLachlan. Freshwater Biology 5: 105-114.

CRUMPTON, W.J. 1977. Notes on occurrence of Odonata in Canterbury and Westland (New Zealand). New Zealand Entomologist 6: 302-304.

CRUMPTON, W.J. 1978. The biology of six South Island ponds. Journal of the Royal Society of New Zealand 8: 179-206.

CRUMPTON, W.J. 1979. Aspects of the biology of Xanthocnemis zealandica and Austrolestes colensonis (Odonataj Zygoptera) at three ponds in the South Island, New Zealand. New Zealand Journal of Zoology 6: 285-297.

CUMMINS, K.W. 1973. Trophic relations of aquatic insects. Annual Review of Entomology 18: 183-206.

CUMMINS, K.W. 1974. Structure and function of stream ecosystems. Bioscience 24: 631-641.

CUMMINS, K.W. & KLUG, M.J. 1979. Feeding ecology of stream invertebrates. Annual Review of Ecology and Systematics 10: 147-172.

CUNNINGHAM, B.T., MOAR, N.T., TORRIE, A.W. & PARR, P.J. 1953. A survey of the western coastal dune lakes of the North Island, New Zealand. Australian Journal of Marine and Freshwater Research 4: 343-386.

DAVIES, R.W. 1969. The production of antisera for detecting specific tric1ad antigens in the gut contents of predators. Gikos 20: 248-260.

DAVIES, R.W. & McCAULEY, V.J. 1970. The effects of preservatives on the regurgitation of gut contents by Chironomidae (Diptera) larvae. Canadian Journal of Zoology 48: 519-522.

DAVIES, R.W. & REYNOLDSON, T.B. 1969. The incidence and intensity of predation on lake-dwelling triclads in the laboratory. Ecology 50: 845-853.

DAVIES, R.W. & REYNOLDSON, T.B. 1971. The incidence and intensity of predation on lake-dwel1ingtric1ads in the field. Journal of Animal Ecology 40: 191-214. 199

DAVIES, R,W" WRONA, F.J, & LINTON, L. 1979. A serological study of prey selection by Helobdella stagnalis (Hirudinoidea). Journal of Animal Ecology 48: 181-194.

DAVIS, S.F. & WINTERBOURN, M.J. 1977. Breakdown and colonization of Nothofagus leaves in a New Zealand stream. Oikos 28: 256-261.

DEACON, K.J. 1979. The seasonality of four Odonata from Mid Canterbury, South Island, New Zealand. Unpublished Ph.D. thesis, Department of Zoology, university of Canterbury, Christchurch, New Zealand. 209p.

DEGEER, C. 1776. Memoires pour servir a l'histoire des Insectes. VI: 1-522. Stockholm.

DEVONPORT, B.F. & WINTE~OURN, M.J. 1976. The relationships of two invertebrate predators in a New Zealand river. Freshwater Biology 6: 167-176.

DISNEY, R.H.L. 1972. Larval Hydroptilidae that prey on Simuliidae in Cameroon. Entomologists Monthly Magazine 108: 84-85.

DUMBLETON, L.J. 1973. An ice chironomid. New Zealand Entomologist 5: 131-133.

EDMONDSON, W.T. & WINBERG, G.G. (Eds) 1971. A manual on methods for the assessment of secondary productivity in fresh waters. International Biological Program Handbook No. 17. 358p.

EDWARD, D.H.D. & COLLESS, D.H. 1968. Some Australian parthenogenetic Chironomidae. Journal of the Austraiian Entomological Society 7: 158-162.

EDWARDS, F.W. 1929. British non-biting midges (Diptera: Chironomidae). Transactions of the Royal Entomological Society, London 77: 279-429.

EHRENBERG, H. 1957. Die Steinfauna der Brandungsufer ostholsteinischer Seen. Archiv fur Hydrobiologie 53: 87-159. ELLIOTT, J.M. 1977. Some methods for the statistical analysis of of benthic invertebrates. 2nd edition. Scientific Publications of the Freshwater Biological Association No. 25. l56p.

ELLIOTT, J.M. & I P.A. 1978. A bibliography of for benthic invertebrates. Occasional Publications of the Freshwater Biological Association No.4. 61p. 200

ENTZ, B. 1947. Qualitative and quantitative studies in the coatings of potamogeton perfoliatus and Myriophyllum spicatum in the Lake Balaton. Archiva biol.Hung. 17: 17-38.

FEENY, P. 1970. Seasonal changes in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillars. Ecology 51: 565-581.

FEENY, P. 1976. Plant apparency and chemical defense. Recent Advances in Phytochemistry 10: 1-40.

FINNEY, D.J. 1946. Field sampling for the estimation of wireworm populations. Biometrics 2: 1-7.

FISCHER, Z. 1966. Food selection and energy transformation in larvae of Lestes sponsa (Odonata) in astatic waters. Internationa1e Vereinigung fur Theoretische und Angewandte Limnologie 16: 600-603.

FISCHER, Z. 1967. Food composition and food preference in larvae of Lestes sponsa (L.) in astatic water environments. Polskie Archiwum Hydrobiologii 14: 59-71.

FISH, G.R. 1966. An artificially maintained trout population in a Northland lake, New Zealand Journal of Science 9: 200-210.

FISHER, R.A., CORBET, A.S. & WILLIAMS, C.B. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology 12: 42-58.

FITTKAU, E.J. 1962. Die Tanypodinae (Diptera: Chironomidae) - Die Tribus Anatopyniini, Macropelopiini und Pentaneurini. Abhandl. Larvalsystematik Insekten 6: 1-453.

FLINT, E.A. 1938. A preliminary study of the phytoplankton in Lake Sarah (New Zealand) . Journal of Ecology 26: 353-358.

FLINT, E.A. 1950. An investigation of the distribution in time and space of the algae of a British water reservoir. Hydrobiologie 2: 217-240.

FLINT, E.A. 1975. Phytoplankton in some New Zealand Lakes. Pp. 162-192 in Jolly, V.H. & Brown, J.M.A. (Eds), New Zealand Lakes. Auckland University Press/Oxford University Press. 388p.

FORSYTH, D.J. 1971. Some New Zealand Chironomidae (Diptera). Journal of the Royal Society of New Zealand 1: 113-144. 201

FORSYTH, D.J. 1975a. The Benthic Fauna. Pp. 281-293 in Jolly, V.H. & Brown, J.M.A. (Eds), New Zealand Lakes. Auckland University Press/Oxford University Press. 388p.

FORSYTH, D.J. 1975b. Description of Kiefferulus opalensis n.sp. (Diptera: Chironomidae). New Zealand Journal of Zoology 2: 215-218.

FORSYTH, D.J. 1978. Benthic macroinvertebrates in seven New Zealand lakes. New Zealand Journal of Marine and Freshwater Research 12: 41-49.

FORSYTH, D.J. 1979. Life stages and taxonomic affinities of Xenochironomus cante'rburyensis (Chironomidae: Diptera). New Zealand Journal of Zoology 6: 467-472.

FORSYTH, D.J. & McCALLUM, LD. 1978a. Xenochironomus canterburyensis (Diptera: Chironomidae) an insectan inquiline commensal of Hyridella menziesi (Mollusca: Lamel1ibranchia). Journal of Zoology, London 186: 331-334.

FORSYTH, D.J. & McCALLUM, I.D. 1978b. Xenochironomus canterburyensis (Diptera: Chironomidae) a commensal of Hyridella menziesi (Larnellibranchia) in Lake Taupo; features of pre-adult life history. New Zealand Journal of Zoology 5: 795-800.

FORSYTH, D.J. & McCOLL, R.H. 1974. The limnology of a thermal lake: Lake Rotowhero, New Zealand: 2. General biology with special reference to the chironomid benthic fauna. Hydrobiologia 44: 91-104.

FREEMAN, P. 1959. A study of the New Zealand Chironomidae (Diptera, Nematocera). Bulletin of the British Museum (Natural History) Entomology 7: 395-437.

FREEMAN, P. 1961. The Chironomidae {Diptera} of Australia. Australian Journal of Zoology 9: 611-737.

FROST, E.W. 1942. River Liffey Survey IV. The fauna of the submerged "mosses" in an acid and an alkaline water. Proceedings of the Royal Irish Academy 13: 293-369.

FULLER, R.L. & STEWART, K.W. 1977. The food habits of stone flies (Plecoptera) in the upper Gunnison River, Colorado. Environmental' Entomology 6: 293-302.

GAGE, M. 1959. On the origin of some lakes in Canterbury. New Zealand Geographer 15: 69-75. 202

GAGE, M. 1977. Glacial geology. Pp. 67-78 in Burrows, C.J. (Ed.), Casso Department of Botany, University of Canterbury, Christchurch, New Zealand. 418p.

GAK, D.Z., GURVICH, V.V., KORELYAKOVA, I.L., KOSTIKOVA, L.E., KONSTANTINOVA, N .A., OLIVARI, G.A., PRIIMACHENKO, A.D., TSEEB, Y.Y., VLADIMIROVA, K.S. & ZIMBALEVSKAYA, L.N. 1972. Productivity of aquatic organism communities of different trophic levels in Kiev Reservoir. Pp. 447-455 in Kajak, Z. & Hillbricht-Ilkowska, A. (Eds) Productivity problems of fresh­ waters, Warsaw-Krakow. Proceedings of the IBP-UNESCO Symposium on Productivity Problems of Freshwaters. 9l8p.

GARNETT, P.A. & HUNT, R.H. 1965. Two techniques for sampling freshwater habitats. Hydrobiologia 16: 114-120.

GERKING, S.D. 1957. A method of sampling the littoral macrofauna and its application. Ecology 38: 219-226.

GIBSON, A.S. 1978. Effects of catchment use on some high-country lakes in Canterbury. UnptiQlished M.Sc. Thesis, Joint Centre for Environmental Sciences, University of Canterbury and Lincoln College, Christchurch, New Zealand. 106p.

GILLESPIE, D.M. & BROWN, C.J.D. 1966. A quantitative sampler for macroinvertebrates associated with aquatic macrophytes. Limnology and Oceanography 11: 404-406.

GLIME, J.M. & CLEMONS, R.M. 1972. Species diversity of stream insects on Fontinalis spp. compared to diversity on artificial substrates. Ecology 53: 458-464.

GODWARD, M.B. 1937. An ecological and taxonomic investigation of the littoral algal flora of Lake Windermere. Journal of Ecology 25: 496-568.

GREENLAND, D.E. 1977. Weather and climate at Casso Pp. 93-116 in Burrows, C.J. (Ed.) Casso Department of Botany, university of Canterbury, Christchurch, New Zealand. 418p.

GREIG, C.A. 1973. A survey of the littoral fauna of two man-made lakes (Lake Aviemore and Lake waitaki). Unpublished M.Sc. Project, University of Canterbury. 52p.

GREIG, C.A. 1976. Ecology of Deleatidium sp. (Ephemeroptera) in Lake Grasmere, Canterbury, with particular reference to its trophic relationships. Unpublished M.Sc. Thesis, University of

Canterbury, Christchurch, New Zealand. l29p. 203

GREEN, J.D. 1974. The limnology of a New Zealand reservoir, with particular reference to the life histories of the copepods Boeckella propinqua Sars and Mesocyclops leuckarti Claus. Internationale Revue der gesamten Hydrobiologie 59: 441-487.

HAMILTON, A.L., SAETHER, O.A. & OLIVER, D.R. 1969. A classification of the nearctic Chironomidae. Fisheries Research Board of Canada Technical Report 124: 1-42.

HANNA, H.M. 1957. A of growth and feeding habits of the larvae of four species of caddisflies. Proceedings of the Royal Entomological Society, London A 32: 139-146.

HARGRAVE, B.T. 1970. The utilisation of benthic microflora by Hyalella azteca (Amphipoda). Journal of Animal Ecology 39: 427-437.

HARROD, J.J. 1964. The distribution of invertebrates on submerged aquatic plants in a chalk stream. Journal of Animal Ecology 33: 335-348.

HICKIN, N.E. 1967. larvae. Hutchinson of London (Hutchinson Co. Ltd.), London. 477p.

HIGLER, L.W.G. 1977. Macrofauna-cenoses on Stratiotes plants in Dutch broads. Rijksinstituut voor Natuurbeheer Verh. 11: 86p.

HIGLER, L.W.G. 1980. Macrofauna in the bottom and littoral vegetation. Pp. 231-239 in Ringelberg, J. (Ed.) Limnological Research in the Maarsseveen Lakes 1975-1980. Department of Aquatic Ecology, University of Amsterdam, The Netherlands. 299p.

HIRVENOJA, M. 1973. Revision der Gattung Cricotopus van der Wulp und ihrer Verwandten (Diptera, Chironomidae). Annales Zoologici Fennici 10: 1-363.

HODGKISS, I.J. & TAl, Y.C. 1976. Studies on Plover Cove Reservoir, Hong Kong, III: A comparison of the species composition of diatom floras on different substrates, and the effects of environmental factors on their growth. Freshwater Biology 6: 287-298.

HOPKINS, C.L. 1976. Estimate of biological production in some stream invertebrates. New Zealand Journal of Marine and Freshwater Research 10: 629-640. 204

HUDSON~ G.V. 1892. An Elementary Manual of New Zealand Entomology. West, Newman & Co., London. 128p.

HUGHES, H.R. 1976. Research into aquatic weeds in New Ze.aland waterways: A review. DSIR Information Series 116: 34p.

HUTTON, F.W. 1902. Additions to the Diptera fauna of New Zealand. Transactions and proceedings of the New Zealand Institute 34: 179-196.

HYNES, H.B.N. 1970. The Ecology of Running Waters. Liverpool University Press. 555p.

IRWIN, J. 1972. New Zealand lakes bathymetric surveys 1965-1970. New Zealand Oceanographic Institute Records 1: 107-126.

JACCARD, P. 1932. Die statistisch-florestische Methode als Grundlage der Pflanzen-sozioligie. Handb. Biol. Arbeitsmeth 5: 165-202.

JAMIN, C.T. 1976. Some aspects of the benthic fauna of Lake Grasmere. B.Sc. Hons. project, Department of Zoology, University of Canterbury, Christchurch, New Zealand. 82p.

JOHNSON, M.G. & BRINKHURST, R.O. 1971. Associations and species diversity in benthic macro invertebrates of Bay of Quinte and Lake ontario. Journal of the Fisheries Research Board of Canada 28: 1683-1697.

JONASSON, P.M. 1978. Edgardo Baldi Memorial Lecture, Zoobenthos of lakes. Internationale Vereinigung fur Theoretische und Angewandte Limnologie 20: 13-37.

KIEFFER, J.J. 1921. Notice sur quelques Chironomides d'Amerique et Nouvelle-Zelande. Annals of the Linnaean Society, Lyon 68: 145-148.

KORINKOVA, J. 1971. Sampling and distribution of animals in submerged vegetation. vestnik Ceskoslovenske Spolecnosti Zoologicke 35: 209-221.

KRANTZ, G.W. & LINDQUIST, E.E. 1979. Evolution of phytophagous mites (Acari). Annual Review of Entomology 24: 121-158.

KRECKER, F.H. 1939. A comparative study of the animal populations of certain submerged aquatic plants. Ecology 20: 553-562.

KUFLIKOWSKI, T. 1974. The phytophilous fauna of the dam reservoir at Goczalkowice. Acta hydrobiol., Krakow 16: 189-207. 205

LADLE, M., BASS, J.A.B. & JENKINS, W.R. 1972. Studies on production and food consumption by the larval Simuliidae (Diptera) of a chalk stream. Hydrobiologia 39: 429-448.

LANGLEY, R. 1968. Practical Statistics. Pan Books Ltd, London. 399p.

LANGTON, P.H. 1980. The genus Psectrocladius Kieffer (Diptera: Chironomidae) in Britain. Entomologists Gazette 31: 75-88.

LAUPY, M. 1977. Relations among the submerged macrovegetation, the quantity of nannoseston and the pond-bottom fauna. vestnik Ceskoslovenske Spolecnosti zoologicke 41: 31-40.

LAWTON, J.H. 1970a. Feeding and food energy assimilation in larvae of the damselfly pyrrhosoma nymphula (SULZ.) (Odonata: Zygoptera) . Journal of Animal Ecology 39: 669-689.

LAWTON, J.H. 1970b. A population study on larvae of the damselfly pyrrhosoma numphula (Sulzer) (Odonata: Zygoptera). Hydrobiologia 36: 33-52.

LEADER, J.P. 1970. Hairs of the Hydroptilidae (Trichoptera). Tane 16: 121-129.

LEADER, J.P. 1972. The New Zealand Hydroptilidae (Trichoptera). Journal of Entomology (B) 41: 191-200.

LEWIS, M.H. 1976. Cladocera. Ppm 47-73 in Chapman, M.A. & Lewis, M.H. (Eds) An Introduction to the Freshwater Crustacea of New Zealand. Collins, London. 261p.

LIDDLE, M.J., HAPPEY-WOOD, C.M. & BUSE, A. 1979. A survey of the biota, environment and use for recreation of twelve lakes in Snowdonia. Biological Journal of the Linnean Society 11: 77-101.

MACAN, T.T. 1949. Survey of a moorland fishpond. Journal of Animal Ecology 18: 160-186.

MACAN, T.T. 1963. Freshwater Ecology. Longman, London. 338p.

MACAN, T.T. 1964. The Odonata of a moorland fishpond. Internationale Revue der gesamten Hydrobiologie 49: 325-360.

MACAN, T.T. 1966. The influence of predation on the fauna of a moorland fishpond. Archiv fur Hydrobiologie 61: 432-452.

MACAN, T.T. 1977~ The fauna in the vegetation of a moorland fishpond as revealed by different methods of collection. Hydrobiologia 55: 3-15. 206

MACAN, T.T. & KITCHING, A. 1972. Some experiments with artificial substrata. Internationale Vereinigung fur Theoretische und Angewandte Limnologie 18: 213-220.

MACAN, T.T. & KITCHING, A. 1976. The colonization of squares of plastic suspended in midwater. Freshwater Biology 6: 33-40.

MACKAY, R.J. & WIGGINS, G.B. 1979. Ecological diversity in Trichoptera. Annual Review of Entomology 24: 185-208.

MACKIE, G.L. & QADRI, S.U. 1971. A quantitative sampler for aquatic phytomacrofauna. Journal of the Fisheries Research Board of Canada 28: 1322-1324.

MAIN, W.de L. 1976. Morphology of Lake Waikaremoana, New Zealand, and reconnaissance of its benthic fauna. New Zealand Journal of Marine and Freshwater Research 10: 597-611.

MALLEY, D.F. & REYNOLDS, J.B. 1979. Sampling strategies and life history of non-insectan freshwater invertebrates. Journal of the Fisheries Research Board of Canada 36: 311-318.

MARPLES, B.J. 1962. An Introduction to Freshwater Life in New Zealand. Whitcombe & Tombs Limited, Christchurch. 160p.

MARSHALL, J.W. 1974. A biological investigation of the Leeston Drain, Canterbury, New Zealand. Unpublished M.Sc. thesis, Zoology Department, university of Canterbury. l36p.

MARTIN, J. 1974. Key to the genera of Australian Tanypodinae larvae (Diptera: Chironomidae). Australian Society for Limnology Newsletter 12: 12-13.

MARTIN, J. 1975. Key to the larvae of Australian genera of Chironomini (Diptera: Chironomidae). Australian Society for Limnology Newsletter 13: 21-22.

MASON, R. 1975. The Macrophytes. Pp. 231-243 in Jolly, V.H. & Brown, J.M.A. (Eds) New Zealand Lakes. Auckland University Press/ Oxford University Press. 388p.

MASON, W.T. 1973. An Introduction to the Identification of Chironomid Larvae. 2nd edition. Divisional Pollution Surveillance, Federal Water Pollution Control Administration, United States Department of the Interior, Cincinnati, Ohio. 90p.

MATLAK, Olga. 1963. The appearance of invertebrates on aquatic plants in fish ponds. Acta Hydrobio1ogica 5: 1-30. 207

McCAULEY, V.J.E. 1975. Two new quantitative samplers for aquatic phytomacrofauna. Hydrobiologia 47: 81-99.

McFARLANE, A.G. 1977. Light trapping of caddisflies at Winchmore Irrigation Research Station, Mid Canterbury, New Zealand (May 1974 - September 1975) (Note). Mauri Ora 5: 123-124.

McLEOD, D. & BURROWS, C.J. 1977. History of the Cass District. Pp. 23-36 in Burrows, C.J. (Ed.) Casso Department of Botany, University of Canterbury, Christchurch. 418p.

McLEOD, D. & McLEOD, I. 1977. The Farming Endeavour. Pp. 37-52 in Burrows, C.J. (Ed.) Casso Department of Botany, University of Canterbury, Christchurch. 418p.

MECOM, J.O. & CUMMINS, K.W. 1964. A preliminary study of the trophic

relationships of the larvae of Brachycentrus american us (Banks) (Trichoptera: Brachycentridae). Transactions of the American Microscopical Society 83: 233-243.

MERRITT, R.W. & CUMMINS, K.W. (Eds) 1978. An Introduction to the Aquatic Insects of North America. Kendall/Hunt Publishing Company, Dubuque, Iowa. 441p.

MERRITT, R.W., CUMMINS, K.W. & RESH, V.H. 1978. Collecting, sampling, and rearing methods for aquatic insects. Pp. 13-31 in Merritt, R.W. & Cummins, K.W. (Eds) An Introduction to the Aquatic Insects of North America. Kendall/Hunt Publishing Company, Dubuque, Iowa. 441p.

MICHAELIS, F.B. 1973. The distribution and life history of Rakiura vernale (Trichoptera, Helicopsychidae). Journal of the Royal Society of New Zealand 3: 295-304.

MICHAELIS, F.B. 1977. Biological features of Pupu Springs. New Zealand Journal of Marine and Freshwater Biology 11: 357-373.

MINTO, M.L. 1977. A sampling device for the invertebrate fauna of aquatic vegetation. Freshwater BioLogy 7: 425-430.

MORGAN, N.C. 1980. Secondary Production. Pp. 247-340 in LeCren, E.D. & Lowe-McConnell, R.H. (Eds) The functioning of freshwater eco­ systems. International Biological Programme 22, Cambridge University Press. 588p.

MOSELY, M.E. 1924. New Zealand Hydroptilidae (Order Trichoptera). Transactions and Proceedings of the New Zealand Institute 55: 671-673. 208

MOSS, B. 1980. Ecology of Freshwaters. Blackwell Scientific Publications, Oxford. 332p.

MRACHEK, R.J. 1966. Macroscopic invertebrates on the higher aquatic plants at Clear Lake, Iowa. Iowa Academy of Science 73: 168- 177.

MULLER-LIEBENAU, I. 1956. Die Besiedlung der Potamogeton-Zone ostholsteinischer Seen. Archiv fur Hydrobiologie 52: 470-606.

MUTTKOWSKI, R.A. & SMITH, G.M. 1929. The food of trout stream insects in Yellowstone National Park. Roosevelt Wildlife Annals 2: 241-263.

NASH, D.A. 1974. Preliminary study of feeding and assimilation of Paratya curvirostris and potamopyrgus antipodarum using radiotracer techniques. Unpublished ph.D. thesis, university of Waikato, New Zealand.

NEEDHAM, P.R. & USINGER, R.L. 1956. Variability in the macrofauna of a single riffle in Prosser Creek, California as indicated by the Surber sampler. Hilgardia 24: 383-409.

NIELSEN, A. 1948. Postembryonic development and biology of the Hydroptilidae. A contribution to the phylogeny of the caddis flies and to the question of the origin of the case-building instinct. Det Kongelige Danske Videnskabernes,Selskab Biologiske Skrifter Bind V, Nr. 1. 200p.

NORRIE, P.H. 1969. The flight activity of Ephemeroptera and Trichoptera in a Waitakere stream. M.Sc. thesis, University of Auckland, New Zealand. 212p.

NURSE, F.R. 1950. Freshwater triclads new to the fauna of New Zealand. Transactions of the Royal Society of New Zealand 78: 410-417.

NYMAN, B. 1971. A comparative study of the bottom fauna on different types of vegetation in Lake Erken. Scripta Limnologica Upsaliensia Collectio 7A 274: 26p.

ODUM, H.T., CANTLON, J.E. & KORNICKER, L.S. 1960. An organizational hierarchy postulate for the interpretation of species individual-distributions, species entropy, ecosystem evolution, and the meaning of a species variety index. Ecology 41: 395- 399.

OLIVER, D.R. 1979. Contribution of life history to taxonomy of aquatic insects. Journal of the Fisheries Research Board of Canada 36: 318-321. 209

OLIVER, D.R., McCLYMONT, D. & ROUSSEL, M.E. 1978. A to some larvae of Chironomidae (Diptera) from the Mackenzie and Porcupine River watersheds. Canadian Fisheries and Marine Service Technical Report 791: iv + 73p.

ORDISH, R.G. 1966. A systematic revision of the New Zealand water beetles (Coleoptera: Dytiscidae). Records of the Dominion Museum 5: 217-264.

PAGAST, F. 1947. Systematik und Verbreitung der urn die Gattung Diamesa gruppierten Chironomiden. Archiv fur Hydrobiologia 41: 435-596.

PATRICK, R. 1959. Bacillariophyceae. Pp. 171-189 in Edmondson, W.T. (Ed.) Freshwater Biology. H.B. Ward & G.C. Whipple, 2nd edition, Wiley & Sons, New York.

PEARLSTONE, P.S.M. 1973. The food of damselfly larvae in Marion Lake, British Columbia. Syesis 6: 33-39.

PENDERGRAST, J.G. & COWLEY, D.R. 1966. An Introduction to New Zealand Freshwater Insects. Collins, Auckland. lOOp.

PENNIKET, J.G. 1969. Ephemeroptera (Mayflies). Pp. 474-477 in Knox, G.A. (Ed.) The Natural History of Canterbury. Reed, Wellington. 620p.

PICKAVANCE, J.R. 1971. The ecology of Lumbriculus variegatus (Muller) (Oligochaeta, Lumbriculidae) in Newfoundland. Canadian Journal of Zoology 49: 337-342.

PIECZYNSKA, E. 1964. Investigations on colonization of new substrates

by N~matodes (Nematoda) and some other periphyton organisms. Polska (A) 12: 185-234.

PIECZYNSKA, E. & SPODIEWSKA, I. 1963. Occurrence and colonisation of periphyton organisms in accordance with the type of substrate. Polska 11: 533-545.

PIELOU, E.C. 1966. The measurement of diversity in different types of biological collections. Journal of Theoretical Biology 13: 131-144.

PINDER, L.C.V. 1978. A key to adult males of British Chironomidae. Vol. 1 The Key 169p., Vol. 2 Illustrations of the Hypopygia (Figs 77-189). Freshwater Biological Association Scientific Publication 37. 210

PRITCHARD, G. 1964. The prey of dragonfly larvae (Odonata: Anisoptera) in ponds in Northern Alberta. Canadian Journal of Zoology 42: 785-800.

RAMSAY, A.J. 1972. Studies on the microorganisms of a freshwater lake. Unpublished Ph.D. thesis, University of Canterbury, Christchurch, New Zealand. 386p.

RAMSAY, A.J. 1974. The use of autoradiography to determine the proportion of bacteria metabolizing in an aquatic habitat. Journal of General Microbiology 80: 363-373.

RAMSAY, A.J. 1976. Heterotrophic bacteria and their relationship with plankton in a New Zealand freshwater lake. New Zealand Journal of Marine and Freshwater Research 10: 77-90.

RESH, V.H. 1976. The biology and immature stages of the caddisfly genus Cerac1ea in Eastern North America (Trichoptera: Leptoceridae). Annals of the Entomological Society of America 69: 1039-1061.

RESH, V.H. 1979. Sampling variability and life history features: basic considerations in the design of aquatic insect studies. Journal of the Fisheries Research Board of Canada 36: 290-311.

RODINA, A.G. & TROSHIN, A.S. 1954. The use of labelled atoms in the study of feeding in aquatic animals. Dok1ady Akademiia nauk SSSR, Moscow 98: 297-300.

ROSINE, N.W. 1955. The distribution of invertebrates on submerged aquatic plant surfaces in Muskee Lake, Colorado. Ecology 36: 308-314.

ROWE, R.J. 1978. Some aspects of the behaviour of xanthocnemis ze1andica larvae. Unpublished B.Sc. Hons. project, Department of Zoology, University of Canterbury, Christchurch. 43p.

SAETHER, O.A. 1969. Some Nearctic Podonominae, Diamesinae, and Orthocladiinae (Diptera: Chironomidae). Fisheries Research Board of Canada Bulletin 170. 154p.

SAETHER, O.A. 1977. Taxonomic studies on Chironomidae: Nanoc1adius, Pseudochironomus, and the Harnischia complex. Bulletin of the Fisheries Research Board of Canada 196: 1-143.

SAETHER, O.A. 1979. Chironomid communities as water quality indicators. Ho1arctic Ecology 2: 65-74. 211

SANDERS, H.L. 1960. Benthic studies in Buzzards III. The structure of the soft bottom community. Limnology and Oceanography 5: 138-153.

SATIJA, G.R. 1974. Structure of the alimentary canal and mouth-parts of Trichoptera larvae with special reference to food and feeding habits. VI. Relationship of alimentary canal, mouth-parts and feeding habits, and phylogeny of the group. Research Bulletin of the Panjab University (Science) 25: 55-70.

SCHLEE, D. 1966. Praparation und Ermittlung von Messwerten an Chironomiden (Diptera). Gewass. Abwass 41/42: 169-193.

SEDELL, J.R. 1972. Feeding rates and food utilization of stream caddisfly larvae of the genus Neophylax (Trichoptera:

Limnephilidae) using 60Co and 14C. Pp. 486-491 in Nelson, D.J. (Ed.) Symposium on radio-ecology; proceedings of the Third National Symposium at Oak Ridge Tennessee, U.S.A. May 8-10, 1971.

SHANNON, C.E. & WEAVER, W. 1949. The Mathematical Theory of Communication. University of Illinois Press, Urbana, Illinois. 117p.

SKOOG, Gunnel. 1978. Influence of natural food items on growth and egg production in brackish water populations of Lymnea peregra and Theodoxus fluviatilis (Mollusca). Oikos 31: 340-348.

SLACK, H.D. 1936. The food of caddisfly larvae (Trichoptera). Journal of Animal Ecology 5: 105-115.

SMYLY, W.J.P. 1952. The Entomostraca of the weeds of a moorland pond. Journal of Animal Ecology 21: 1-11.

SOONS, J.M. 1977. The geomorphology of the Cass District. . 79-92 in Burrows, C.J. (Ed.) Casso Department of Botany, University of Canterbury, Christchurch. 418p.

SOROKIN, J.I. 1966. Carbon-14 method in the study of the nutrition of aquatic animals. Internationale Revue der gesamten Hydrobiologie und Hydrographie 51: 209-224.

SOSZKA, G.J. 1975a. The invertebrates on submerged macrophytes in three Masurian lakes. Ekologia polska 23: 371-391.

SOSZKA, G.J. 1975b. Ecological relations between invertebrates and submerged macrophytes in the lake littoral. Ekologia Polska 23: 393-415. 212

STADDON, B.W. & GRIFFITHS, D. 1967. Some observations on the food of Aeshna juncea (L.) nymphs (Odonata) with particular reference to Corixidae (Hemiptera). Entomologists Monthly Magazine 103: 226-230.

STARK, J.D. 1980. A cylinder-sampler for collecting the invertebrate fauna from submerged aquatic vegetation. Mauri Ora 8: 45-53.

STARK, J.D. in press. Keys to (a) larval Chironomidae (non-biting midges) and (b) larval Hydroptilidae in winterbourn, M.J. & Gregson, K.L.D., Guide to the Aquatic Insects of New Zealand. Bulletin of the Entomological Society of New Zealand.

STARK, J.D., FORDYCE, R.E. & WINTERBOURN, M.J. 1976. An ecological survey of the hot springs area, Hurunui River, Canterbury, New Zealand. Mauri Ora 4: 35-52.

STEPHENSON, W., WILLIAMS, W.T. & COOK, S. 1974. The macrobenthos of soft bottoms in Southern Moreton Bay (south of Peel Island). Memoirs of the Queensland Museum 17: 73-124.

STEPHENSON, W., WILLIAMS, W.T. & LANCE, G.N. 1970. The macrobenthos of Moreton Bay. Ecological Monographs 40: 459-494.

STOUT, V.M. 1953a. Eylais waikawae n.sp. (Hydracarina) and some features of its life history and anatomy. Transactions and proceedings of the Royal society of New Zealand 81: 389-416.

STOUT, V.M. 1953b. New species of Hydracarina, with a description of the life history of two. Transactions of the Royal Society of New Zealand 81: 417-466.

STOUT, V.M. 1969a. Lakes in the mountain of Canterbury, New Zealand. Internationale Vereinigung fur Theoretische und Angewandte Limnologie 17: 404-413.

STOUT, V.M. 1969b. Life in lakes and ponds. Pp. 458-470 in Knox, G.A. (Ed.) The Natural History of Canterbury. Reed, Wellington. 620p.

STOUT, V.M. 1970. The invertebrate animals of New Zealand inland waters, and factors influencing their distribution. Proceedings of the New Zealand water Conference Part 1: 11.1- 11.24.

STOUT, V.M. 1972. Plankton composition in relation to nutrient inflow in a small New Zealand lake. Internationale Vereinigung fur Theoretische und Angewandte Limnologie 18: 605-612. 213

STOUT, V.M. 1975a. Lake Grasmere: a study of a small South Island lake. Pp. 110-122 in Jolly, V.H. & Brown, J.M.A. (Eds) New Zealand Lakes. Auckland University Press/Oxford University Press. 388p.

STOUT, V.M. 1975b. The limnology. Pp. 405-458 in Kuschel, G. (Ed.) Biogeography and Ecology in New Zealand. Dr W. Junk, B.V. Publishers, The Hague.

STOUT, V.M. 1976. Class Arachnida. Pp. 199-214 in Chapman, M.A. & Lewis, M.H. (Eds) An Introduction to the Freshwater Crustacea of New Zealand. Collins, Auckland. 261p.

STOUT, V.M. 1977. Biology of the fauna of lakes and tarns. Pp. 291- 310 in Burrows, C.J. (Ed.) Cass, History and Science in the Cass District, Canterbury, New Zealand. Department of Botany, University of Canterbury. 418p.

STUBE, M. 1958. The fauna of a regulated lake. Report from the Institute of Freshwater Research, Drottningholm39: 162-224.

SUBLETTE, J.E. & WIRTH, W.W. 1980. The Chironomidae and Ceratopogonidae (Diptera) of the subantarctic islands of New Zealand. New Zealand Journal of Zoology 7: 299-378.

TARWID, M. 1969. Analysis of the contents of the alimentary tract of predatory Pelopiinae larvae (Chironomidae). Ekologia Polska, Series A 17: 125-131.

TAYLOR, B.J.R. 1965. The analysis of polymodal frequency distributions. Journal of Animal Ecology 34: 445-452.

TAYLOR, C.C. 1953. Nature of variability in trawl catches. Fishery Bulletin, united States Fish & wildlife Service 54: 145-166.

THOMPSON, D.J. 1978a. The natural prey of larvae of the damselfly, Ischnura elegans (Odonata: Zygoptera). Freshwater Biology 8: 377-384.

THOMPSON, D.J. 1978b. Prey size selection by larvae of the damselfly Ischnura elegans (Odonata). Journal of Animal Ecology 47: 769- 785.

THUT, R.N. 1969. Feeding habits of larvae of seven Rhyacophila species with notes on other life-history features. Annals of the American Entomological Society 62: 894-898. 214

WHITTAKER, R.H. 1975. communities and Ecosystems. 2nd edition. Macmillan, New York. 385p.

WHITTAKER, R.H. & FAIRBANKS, C.W. 1958. A study of plankton and copepod communities in the Columbia Basin, southeastern washington. Ecology 39: 46-65.

WIGGINS, G.B. 1977. Larvae of the North American genera (Trichoptera). University of Toronto Press, Toronto. 401p.

WIGGINS, G.B. 1978. Chapter 14 in Merritt, R.W. & Cummins, K.W. (Eds)

An Introduction to the Aquatic Insects of North America. Kendall/Hunt, Dubuque, Oiwa, U.S.A. 441p.

WIGGINS, G.B. & MACKAY, R.J. 1978. Some between systematics and.trophic ecology in Nearctic insects, with special reference to Trichoptera. Ecology 59: 1211-1220.

WILSON, K.E. 1980. Studies on the distribution of shell phenotypes and predation on Potamopyrgus antipodarum Unpublished B.Sc. (Hons) project, Zoology University of Canterbury, Christchurch, New Zealand. 56p.

WINTER, J.W. 1964. A survey of Lake Roxburgh, a recent hydro-electric dam. Proceedings of the New Zealand Ecological Society 11: 16-23.

WINTERBOURN, M.J. 1966. The ecology and life history of Zelandoperla maculata (Hare) and Aucklandobius trivacuatus (Tillyard). New Zealand Journal of Science and Technology 9: 312-323.

WINTERBOURN, M.J. 1968. The faunas of thermal waters in New Zealand. Tuatara 16: 111-122.

WINTERBOURN, M.J. 1970a. The New Zealand of Potamopyrgus (Gastropoda: Hydrobiidae). 10: 283-321.

WINTERBOURN, M.J. 1970b. Population studies on the New Zealand freshwater gastropod Potamopyrgus antipodarum (Gray). Proceedings of the Malacological , London 39: 139-149.

WINTERBOURN, M.J. 1971a. The life histories and trophic relationships of the Trichoptera of Marion Lake, British Columbia. Canadian Journal of Zoology 49: 623-635.

WINTERBOURN, M.J. 1971b. An ecological study of Banksiola crotchi Banks (Trichoptera, Phryganeidae) in Marion Lake, British Columbia. Canadian Journal of Zoology 49: 637-645. 215

WINTERBOURN, M.J. 1973. A guide to the freshwater Mollusca of New Zealand. Tuatara 20: 141-159.

WINTERBOURN, M.J. 1974. The life histories, trophic relations and production of Stenoperla prasina (Plecoptera) and Deleatidium sp. (Ephemeroptera) in a New Zealand river. Freshwater Biology 4: 507-524.

WINTERBOURN, M.J. 1978. The food and occurrence of larval Rhyacophilidae and Polycentropodidae in two New Zealand rivers. Pp. 55-66 in Chrichton, M.I. (Ed.) Proceedings of the Second International Symposium on Trichoptera. Junk, The Hague. 359p.

WINTERBOURN, M.J. 1980. The freshwater insects of Australasia and their affinities. Palaeogeography, Palaeoclimatology, Palaeoecology 31: 235-249.

WINTERBOURN, M.J. & DAVIS, S.F. 1976. Ecological role of Zelandopsyche ingens (Trichoptera: Oeconesidae) in a beech forest stream ecosystem. Australian Journal of Marine & Freshwater Research 27: 197-215.

WINTERBOURN, M.J. & GREGSON, K.L.D. in press. Guide to the aquatic insects of New Zealand. Bulletin of the Entomological Society of New Zealand.

WINTERBOURN, M.J. & LEWIS, M.H. 1975. Littoral fauna. Pp. 271-280 in Jolly, V.H. & Brown, J.M.A. (Eds) New Zealand Lakes. Auckland University Press/Oxford University Press. 388p.

WINTERBOURN, M.J. & STARK, A.W. 1978. A biological survey of the Kaiapoi River following fellmongery effluent treatment. Mauri Ora 6: 11-21.

WISE, K.A.J. 1972. Trichoptera of the Auckland Islands. Records of the Auckland Institute and Museum 9: 253-267.

WISE, K.A.J. 1973. A list and bibliography of the aquatic and water- associated insects of New Zealand. Records of the Auckland Institute and Museum 10: 143-187.

WISE, K.A.J. 1978. Further records of subantarctic Trichoptera. Records of the Auckland Institute and Museum 15: 111-113.

YOUNG, E.C. 1962. The Corixidae and Notonectidae (Hemiptera; Heteroptera) of New Zealand. Records of the Canterbury Museum 1: 327-374. 216

TIM}1S, B.V. 1980. The macrobenthos of Lakes Rotoroa and Rotoiti, South Island, New Zealand, with special reference to the influence of allochthonous organic detritus. Archiv fur Hydrobio1ogie 90: 182-196.

TOWNS, D.R. 1976. Dynamics of benthic invertebrate communities in a northern New Zealand kauri forest stream ecosystem. Unpublished Ph.D. thesis, University of Auckland, Auckland, New Zealand. l50p.

WALKER, A.K. & CROSBY, T.K. 1979. The preparation and curation of insects. New Zealand Department of Scientific and Industrial Research and the Entomological Society of New Zealand. DSIR Information Series 130: SSp.

WATT, J.C. 1979. Biosystematics: the neglected science. New Zealand Science Review 36: 68-72.

WEBER, C.I. 1971. A guide to the common diatoms at water pollution surveillance system stations. U.S. Environmental Protection Agency, Cincinatic, Ohio. 101p.

WELCH, P.S. 1916. Contributions to the biology of certain aquatic Lepidoptera. Annals of the Entomological Society of America 9: 159-186.

WELCH, P.S. 1924. Observations on the early larval activities of Nymphula macula1is Clemens (Lepidoptera). Annals of the Entomological Society of America 17: 395-402.

WELCH, P.S. 1948. Limnological Methods. Philadelphia, Blakiston Press. 38lp.

WESTLAKE, D.F. 1975. Macrophytes. Pp. 106-128 in Whitton, B.A. (Ed.) River Ecology. Blackwell Scientific Publications, Oxford. 725p.

WETZEL, R.G. 1975. Limnology. W.B. Saunders Co., London. 743p.

WETZEL, R.G. & ALLEN, H.L. 1972. Functions and interactions of dissolved organic matter and the littoral zone in lake metabolism and eutrophication. Pp. 333-347 in Kajak, Z. & Hillbricht-Ilkowska, A. (Eds) Productivity problems of freshwaters. Warsaw & Krakow: Polish Scientific Publishers. 9l8p.

WHITEHEAD, H. 1935. An ecological study of the invertebrate fauna of a chalk stream near Great Driffield, Yorkshire. Journal of Animal Ecology 4: 58-78. 217

YOUNG, E.C. 1970. Seasonal changes in populations of Corixidae and Notonectidae (Hemiptera: Heteroptera) in New Zealand. Transactions of the Royal Society of New Zealand 12: 113-130.

ApPEIndix 1. Number. of invertebrates per sample (- 0.006 m2) collected from Lake Grasmere during the pilot survey (14 A~)ril 197G) using the cylinder-sampler. (E • Elodea can~densls, I - Isoetes alpinu8, R - ~nunculus {lui tans)

- .. Sa.mpll1ig: area N o R T II SOU T II E A S T II E S T 2 1 Lo>.ke d~pth ha) 2 2 2 1 1 1 1 1 1 1 2 2 2 2 2 1 1 1 0.6 0.5 0.7 0.6 2 2 2 2 1 1 1 1 1 E E E E E E E R R R R i'lacrophyte Ii. Ii. E E E E E I 1 1 I I I E E E E E E E I I Dry wt 0.87 2.19 2.09 1. 62 1.13 1.16 1.82 0.90 0.40 0.S3 0.95 2.03 2.02 :':.23 1.96 2.74 1.06 1.45 O.Sl 0.76 0.26 0.35 0.23 0.16 0,59 0.62 0.97 1. 25 0,51 0.57 mflcrolJhyte (9l 0.32 1. )0 1. 25

Inv.:rtebrate taxon

Chlorohydc.& 3 10 2~ 1 6 1 100 3 31 23 12 2 vlridissiQcl 7 44 1 11 37 1 5 1 12 4 2 Curd pin9uis 2 1 1 1 8 1 4 1 3 1 4 1 Plu~tella repens P Ii P P Ii P Ii P Ii P P Ii Ii Ii P P Ii' P p P Ii P C~eeogascer sp. 5 36 19 28 (, 14 13 6 6 11 6 1 36 13 1 Other OLlCOCHAETA 1 3 4 61 12 130 1 57 1 6 1 209 2 3 20 1 5 NEMATOOA 1 1 1 1 1 1 Glos~lphonla sp. 1 1 1 wudenL CLAOOCERA 45 17 a 62 7 26 27 6 17 57 6 5 24 10 10 3 2 1 4 27 63 9 2 51 28 27 16 7 7 20 OSTRACODA 2 1 2 1 2 7 '9 6 '3 19 96 2 8 2 5 4 Eucyclops serrulatu$ 5 13 7 5 4 15 1 3 12 1 5 10 3 4 3 36 135 148 1 9 2 12 2 3 xdnthocnemls zea,.t.andicd 1 1 11 3 5 2 1 1 1 1 1 1 1 1 1 1 1 ;< Diaprepocoris z£>dlandi.de 1 1 Aneiporus scrigo$ulu$ 1 1 1 1 Oxyethica albiceps 1 1 1 P.3roxyethira hendersoni 1 2 21 13 10 12 1 4 10 14 22 4 4 2 1 3 2 2 1 1 2 2 4 1 1 6 P. c>llyardi - 2 3 9 3 3 prec:ase Ii'iDROP'l'ILlDAE . 1 2 52 44 9 19 6 6 27 66 3 15 7 4 4 5 2 6 ~6 16 1 1 12 2 33 10 9 24 ~2 Triplectides cep~lotes 1 1 1 Hudsonema amabilis 4 1 1 8 13 27 NymFi:ul .. nitens 1 1 1 2 1 2 1 1 CIIIRONOMlDAE 2 2 1 6 18 7 1 1 7 1 2 e 8 I) 5 2 5 10 H'iDRACAIUNA a 32 7 31 16 10 39 9 11 9 1 66 33 30 49 66 37 43 17 14 1 7 33 64 13 15 5 4 4 4 7 20 48 Gyc.uulu.s corjnn£ 138 367 292 219 6S 45 266 51 3 3 ·13 75 26 30 13 31 7 3 16 1 3 I) 5 10 37 42 45 58 20 38 21 50 Potamapyrgus .antipodarwn 168 506 373 83a 834 455 627 230 116 460 317 294 1012 2360 665 1893 794 1540 70S 604 144 739 1603 1435 92 140 444 130 506 379 :lUi 283 1602 Spl'Ji.l.eriun:! novoaezelandl,ae 1 11 17 9 7 11 1 12 1 1 1 tv tv o Appendix 2.1 Numbers of invertebrates per sample (; 0.008 m2 ) of Elodea canadensis collected from tha northern sampling araa (water depth NE

Date 219/76 2/11/76 2112/76 20/1/771 2/3/77 8/4/77 10/5/77 20/6/77 I 1317/77 I 3110/77 Dry wt 0.58 3.03 0.92 0.52 0.68 0.35 0.27 0.22 0.18 0.26 0.28 0.20.0.34 0.22 0.45 1.26 1.04 1.29 1. 01 1.61 0.30 0.720.37 0.66 : 0.11 0.32 0.35 0.40 0.34 O.4~ (g) macrophyte i Invertebrate taxon

Chlorohydra viridissima 3 1 1 11 19 4 35 48 46 4 4 9 17 37 39 5 37 14 11 3 8 3 Cura pinguis 1 1 1 3 1 2 1 1 1 1 Plumatella repens p p P P P P P P P P P P P P P P P I P Chaerogaster sp. 10 1 13 1 1 13 19 6 5 1 1 11 "\ 2 4 4 7 5 1 5 9 5 8 5 1 8 other OLIGOCHAETA 3 114 70 57 10 17 5 33 27 8 21 1 33 17 11 94 4 1 9 4 2 1 4 1 1 Glossiphonia sp. 1 5 1 1 Placobdella maorica 1 Bosmina meridionalis 1 2 44 4 167 195 55 29 2 Graproleberis 8 23 7 10 2 5 1 6 20 1 1 cestudinaria C]lydoru$ sphaer iells 2 1 ceriodaphnia dubia 1 1 1 ! unident. CLADOCERA 7 1 8 6 61 72 52 53 7 OSTRACODA 1 i Eucyclops serrulatus 2 3 1 12 11 1 28 6 24 54 127 109 94 25 27 402 316 409 45 77 49 9 39 10 5 5 8 27 17 18 procordulia grayi 1 Xan thocnemis 2 1 zealandica Diaprepocoris (Al 1 zealandiae (Ll 1 Sigara arguta (Al 1 I 1 1 2 7 2 2 6 4 5 3 i Paroxyer:.hira (L) 3 7 1 1 2 2 1 1 hendersoni (1'1 2 1 : paroxyer:.hira 1 1 (Ll 1 1 cillyardi (1') pracase HYDROPTILIDAE 1 3 1 2 1 4 B 3 3 3 2 1 5 1 1 2 Tr"plect.i.des (L) 1 1 1 1 cephalotes Hudsonema amabilis (L) 1 CHI RONOMIDAE 9 1 1 1 1 3 1 1 5 5 2 4 2 2 24 8 3 4 4 Piona uncata exigua 3 10 4 7 3 5 3 6 2 1 22 17 19 B 14 10 6 8 15 4 5 10 3 5 Hydrozetes lemnae 4 3 3 6 5 8 4 4 2 4 8 1 4 1 16 2 3 2 3 3 1 4 1 1. TrimGlaconothrus (Al 1 5 8 2 4 5 5 1 1 1 novus (Ll 10 4 Cyraulus corinna 27 54 5 54 34 27 42 42 8 34 36 20 41 34 79 106 113 83 107 69 18 15 2 37 6 21 14 71 Pocamopyrgus 79 249 89 159 172 106 74 74 57 63 69 32 118 38 218 540 411 452 350 495 135 122 83 64 58 126 39 130 171 304 . an tipoda rum I NEMATODA 1 1 1 1 14 1 4 3 4 I 1

(Al Adult, (L) Larva, (P) ~ Pupa. For Plumar:.ella repens, P present. Appendix 2.2 Numbers of invertebrates per sample (- 0.008 m2) of IsaeCes alpinus collected from the southern sampling area (water depth 0- 1 m), SI September 1976 - OCtober 1977.

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Dry wt ll!Acrophyte (9) 0.28 0.28 0.39 0.19 0.20 0.11 0.26 0.17 0.19 0.03 0.12 0.03 Invertebrate Taxon

Chlorohydra viridissima 2 2 11 19 10 7 2 1 S 1 Cura pinguis 1 1 2 2 '\I 1 1 1 Plumatella repens p P P P P C1)aetogaster 51'. 2 13 15 3 '\I 2 3 12 4 other OLIGOCHAETA 3 1 64 101 2 5 7 15 1 2 Glossiphonia sp. 1 Bosmina meridionalis 12 3 Alona gutcata 11 24 52 10 17 GrapColeberis testudinaria 13 3 Chydorus sphaericus 2 3 1 3 uniden t. • CI.J\.DOCERA 1 13 24 /I 14 OSTRACODA 1 4 1 1 Eucyclops serrulat.us 4 4 3 7 14 5 8 1 10 16 12 20 Deleatid.J.um sp. 1 Zelandobius furcillatus 1 Diaprepocoris zealandiae 1 (Al 1 1 Oxyethira a!biceps (Ll 3 Paraxyethira hendersoni (L) 1 10 8 4 9 '2 9 4 1 (P) 1 1 PdroJ(yethira tl11 yardJ. ILl 1 (P) precase HYDROPTILIOAE 2 1 23 11 7 11 Triplectides cephalotes (Ll 1 2 Oececis unicolor (Ll 2 Oecetis iei (Ll 1 Nymphula nitens 1 CHlRONOMIDAE ILl 2 2 9 2 30 3 6 20 22 37 4 2 (P) 1 Piona uncata ~xigua 2 3 8 6 1 3 HydrazeCes lemnae 1 11 26 28 20 12 6 7 6 .:I 4 1 Trimalaconothrus novus (AI 1 1 (Ll 1 1 Gyraulus corinna 1 6 8 6 20 28 28 57 43 10 10 3 Pocamopyrgus antjpod~rum 69 62 69 85 17 129 128 241 133 368 Sphaerium novaezelandiae 42 23 1 3 11 NEMATOOl\. 1 i I 12 6 I 12 1 Appendix 2.3 Numbers of invertebrates per sample (= 0.008 m2) of Myriophyllum propinquum collected from the southern sampling area s (water depth = 0 - 1 m), September 1976 - October 1977.

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 1317/77 I 3/10/77 Dry wt macrophyte (g) 0.67 0.37 0.47 0.24 0.27 0.36 0.53 0.59 0.18 Invertebrate taxon I i

Chlorohydra viridissima 9 4 5 6 4 Cura pinguis 4 PlumaCella repens P P P P P Chaetogaster sp. 11 12 6 7 7 4 other OLlGOCHAETA 3 3 2 9 12 4 3 Bosmina meridionalis 3 1 Alona guttaea 3 31 2 Graptoleberis testudinaria 9 5 22 1 Chydorus sphaericus 12 33 Ilyocrypcus sordidus 2 Neochrix armata 1 uniden t. CLADOCERA 21 1 19 1 OSTRACODA 6 14 12 Eucyclops serrulatus 14 8 3 1 1 32 26 31 1 Zelandobius furcillatus 1 Diaprepocoris zealandiae (Al (Ll 1 1 Sigara arguea (Ll 1 ParoxyeChira hendersoni (L) 7 19 3 11 2 4 5 2 (P) 1 1 1 precase HYDROPTILIDAE 5 1 5 20 10 3 4 1 Pycnocencrodes aureola (Ll 1 6 1 Triplectides cephalotes (Ll 5 2 3 1 T. obsoleta (Ll 1 Hudsonema amabilis (Ll 1 1 2 Nymphula nitens (Ll 1 1 1 2 2 1 (P) 1 CHIRONOMIDAE (Ll 4 1 12 22 8 8 78 55 8 (F) 1 1 Piona uncata exigua 5 1 4 5 5 Hydrozetes leronae 18 17 2 ':3 15 12 3 Trimalaconochrus nevus (Al 1 3 2 2 3 4 (Ll 1 1 8 5 Gyraulus corinna 14 24 27 5 11 101 12 5 Pocamopyrgus ancipodarum 43 79 95 9 109 1699 166 67 97 sphaerl novaezelandiae """,.,,.. "'..,'"' 3 1 I I IIppendix 2.4 N~r. of invertebrates per sa.ple (m 0.008 m'2) of Elodea canaden.i. collected from the eastern sampling are~ (water depth E 1 m), September 1976 - October 1977.

Date 2/9/76 2/11/76 2/12/76 i 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Dry \art I'lIoacrophyte {g} 0.55 1.09 0.64 0.27 0.84 0.41 lnvcrtebr~te taxon

Chloroh~drd viridissima 2 51 191 12 Cur,;; pinguis 1 10 1 l'll.lmatell.!! repens p p p p Chdetogaster sp. 1 5 other OLlGOCHAETA 15 3 6 Glossipilonia Spa 1 Alena gl.ltl:ollt,a 4 Graptoleberis testudinaria 1 unident. CL.!\OOCEAA 9 4 1 OSTRACODA 4 Euc~clops serrulatus 52 3 1 54 86 I Diaprepocoris zealandiae (AI 1 Antiporus strigosulus (A) I ILl 1 Parox~ethi.ra hendersoni ILl 6 P. I ti11!lardi ILl 3 11 2 9 precase HYDROPTILlDAE 1 5 3 1 P~cnoc:elltrodes aureolQl. (L) 1 I Hudsonema amabilis (Ll 1 HYliIphula nitens '2 CHIRONO.IUDAE 2 1 '2 6 Piona uncata exigua 1 17 2 10 4 20 H!ldrozetes lemnae 20 '24 '2 8 18 Trimalaconothl'us novus (Al 3 3 1 I ILl 1 1 Gyraulus corinna 4 1 2 6 9 PotalilOpyrgus antipodarum 109 712 81 135 827 526 SphaeriUlil novaezelandiae 1 13 NEMATODA 3 l 2 Appendix 2.5 Numbers of invertebrates per sample (= 0.008 m ) of Elodea canadensis collected from the eastern sampling area (water depth = 2 m), September 1976 - October 1977. EE2m

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Ory wt 0.88 0.92 0.89 0.74 0.81 0.74 0.44 0.50 0.79 0.47 0.77 0.55 0.25 0.53 0.30 1.10 0.72 0.68 0.63 0.63 0.50 0.62 0.69 0.24 0.16 0.49 0.38 0.40 macrophyte (g) Invertebrate taxon

Chlorohydra vir idissima 1 1 5 8 13 9 58 76 127 13 64 18 209 103 85 27 7 53 1 13 2 5 13 69 8 Curd pinguis 5 2 1 3 3 1 3 1 3 Plumatella repens P P P P P P P P P P P P P P P P P P P P P Chaetogaster sp. 1 2 2 2 3 1 1 1 4 4 1 3 2 1 other OLlGOCHAETA 53 1 7 6 4 1 13 2 3 3 10 10 9 2 18 5 8 19 30 1 2 Glossiphonia sp. 1 1 Bosmina meridionalis 1 10 Alona guttat.a 1 Graptoleberis testudinaria 1 1 1 1 Chydorus sphaericus 3 unident. CLADOCERA 13 14 18 6 2 3 2 3 11 8 8 6 OSTRACODA 1 2 2 4 4 2 1 5 1 10 13 7 2 1 1 3 16 Eucyclops serrulatus 3 4 11 5 1 8 3 11 23 3 12 24 7 113 238 71 41 3 28 14 45 10 11 11 11 2 Diaprepocoris (A) 1 4 4 zealandiae (L) 1 1 Paroxyethira (L) 1 1 1 2 1 1 1 1 hendersoni (P) P. tillyardi (L) 3 2 2 1 5 1 1 5 1 1 6 4 3 1 6 1 3 6 2 7 precase HYOROPTILlDAE 1 1 1 1 1 3 2 2 pycnocentrodes aureola (L) 1 2 Triplectides obsoleta (L) 1 Hudsonerna arnabilis (L) 1 2 1 2 1 2 2 1 12 1 Nymphula nitens 1 1 1 1 CHIRONOHIDAE 5 1 1 1 2 3 2 4 3 1 5 1 3 12 2 1 8 1 Arrenurus sp_ 1 " Piona uncata exigua 5 12 8 9 4 5 2 17 32 8 7 15 7 4 7 3 23 4 4 5 2 6 8 4 9 13 Hydrozetes lemnae (A) 2 21 10 13 6 9 7 12 14 4 8 4 8 3 1 4 3 4 1 2 1 1 5 1 8 (L) 1 Tr imalaconothrus (A) 4 1 2 1 novus (L) 1 Gyraulus corinna 5 4 2 3 12 1 9 13 3 9 12 12 21 15 35 8 13 7 2 9 3 5 Physastra variabilis 1 1 1 1 2 1 potamopyrgus antipodarum 385 359 145 372 175 370 130 85 96 114 137 147 116 565 392 503 681 304 515 346 407 376 329 195 178 148 108 245 Sphaerium novaezelandiae 8 2 1 2 2 3 6 7 4 2 6 NEMATODA 1 1 1 2 1 1 1 1 Appendix 2.6 Numbers of invertebrates per s&mple (-0.008 m2) of Isoetes alpinus collected from the eastern sampling area (water depth - 0-1 m), September 1976 - October 1977. EI - Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Dry yt m.acrophyte (q) 0.42 0.50 0.25 0.19 0.14 0.06 0.05 0.10 0.12 0.20 0.03 0.05 0.05

Invertebrate taxon I

Chloroh~dra viridissima 1 7 22 12 2 :;: 1 Cura pinguis 2 , 3 1 1 1 Chdetogaster sp. 2 1 J other PLIGOCHAETA :2 2 1 2 Sasmina meridionali. 5 Alona gutta!:a :2 8 1 Graptoleberis testudinari~ 6 15 16 Chydorus sphderic13s 1 1,JJlident. CLAOOCERA IS 2 3 5 1 13 Gomphoc~there duff.! 2 2 1 other OSTRACODA 1 1 3 3 21 1 Eucyclops serrulatus 6 34 2 13 42 23 3 103 54 86 114 49 Deleatidium sp. 1 Zelandobius furcillatus 4 Diaprepocoris zealandias 1.1\) ILl 1 I An!:iporus scrigosulus (11) 1 CLl paroxyethirill bendersoni. iLl 2 1 1 Ii?) 1 P. tillyardi CL) 2 1 Ii?) 1 precase HYDROPTILIDAE 1 S 2 27 7 3 3 1 1 3 ! Pycnocenerodes aureola (L) 1 1 ii? ) Hudsonama .mabilis (L) 1 1 1 l CF) N~mphuldl nit

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 B/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Dry wt rnacrophyte (g) 0.57 0.44 0.4B 0.33 0.38 0.26 0.19 0.41 0.45 0.26 Invertebrate taxon i

Chlorohydra viridissima 1 5 11 109 21 13 2 1 Cura pinguis 1 1 3 1 Plumatella repens p p p p Chaetogaster sp. 9 1 9 ] 3 other OLlGOCHAETA 2 1 Graptoleberis testudinaria 2 Chydorus sphaericus 1 2 unident. CLADOCERA 6 3 31 3 OSTRACODA· 1 Eucyclops serrulatus 5 24 28 1 1 14 9 5 Sigara arguta (A) 1 Paroxyethira hendersoni (Ll 2 3 3 2 (P) 1 P. tillyardi (L) 2 2 2 1 precase HYDROPTILIDAE 15 25 6 1 4 9 1 Hudsonema a~~bilis (L) 2 1 1 1 5 Triplectides =ephalotes (Ll (P) 1 tJymphula nitens (L) 1 1 1 3 2 2 (P) 1 CHIRONOMIDAE 2 4 10 2 6 1 Piona uncata exigua 5 1 5 2 13 5 2 Hydrozetes lemnae 1 4 24 29 14 7 1 2 Trimalaconothrus novus (A) 1 1 2 1 (L) Gyraulus corinna 1 2 2 4 1 2 1 Physastra variabilis 1 Potamopyrgus antipodarum 59 45 313 155 113 118 241 49 19 27 NEMATODA 1

i 2 Appendix 2.8 Numbers of invertebrates per sample (= 0.008 m ) of Ranunculus fluitans collected from the eastern sampling area (water depth = 1 - 2 m) •. September 1976 - October 1977.

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77

Dry 10"1:. macrophyte (g) 0.66 0.47 0.26 0.24 0.40 0.25 Invertebrate taxon I

Chiorohydra viridissima 1 5 6 Cura pinguis 1 Plumatella repens p p Chaetogaster sp. 1 1 other OLlGOCHAETA 2 1 Graptoleberis testudinaria 7 unident. CLADOCERA 2 1 OSTRACODA 2 Eucyclops sexrulatus 9 1 9 9 2 19 Sigara arguta (A) . 1 Paroxyethird hendersoni (L) (P) 2 P. t:illyardi (Ll 1 1 precase HYDROPTILlDAE 8 11 8 Hudsonema amabilis ILl 2 1 1 1 CHI RONOMIOAE 2 1 9 Piona uncata exigua 1 1 2 2 12 4 Hydrozetes lemnae 3 8 19 23 1 Trimalaconothrus novus (A) 1 2 (Ll 1 Gyraulus corinna 1 1 1 6 Pot:amopyrgus ant:ipodarum 12 26 143 84 70 392 Sphaerium novaezelandiae 1 NEMATODA 1 I Appendix 2.9 Numbers of invertebrates per sample (z O.OOB m2 ) of Elodea canadensis collected from the western sampling area (water depth E 1 m), September 1976 - October 1977. N N E 00

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 B/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Dry wt macrophyte (91 0.63 0.92 0.66 0.41 0.39 0.65 0.5B 0.61 0.27

Invertebrate taxon I

Chlorohydra viridissi~ 7 4 14 IB 12 Cura pinguis 4 1 Plumatella repens P P P l? P P Chaetogaster sp. 67 9 11 130 17 12 other OLlGOCHAETA 3 71 21 3 3 6 6 201 Bosmina meridionalis 3 10 Alona guttata 1 12 Graptoleberis testudinaria 1 65 67 5 Chydorus sphaericus 5 1J.\"lident. CLADOCERA 4 126 11 OSTRACODA 7 13 15 2 2 Eucyclops serrulatus 3 3 2 12 17 25 130 140 9 DeleatidiWll sp. 1 Xanthocnemis zealandica 3 2 1 1 Oxyethira albiceps (Ll - 1 1 Paroxyethira hendersoni (Ll 16 9 3 5 4 5 2 (p) 25 1 P. tillyardi (L) 2 precase HYDROPTILlDAE 4 1 1 6 10 1 Pycnocentrodes aureola (Ll 1 Triplectides cepnalotes (Ll 1 1 1 Oecetis unicolor (Ll 1 Hudsonema amabilis (L) 1 CHIRONOMIDAE 2 2 2 1 3 76 6 11 Arrenurus sp. 1 Piona uncata exigua :2 2 1 11 4 1 1 Hydrozetes lemnae 1 9 6 2 1 2 Trimalaconotbrus novus (A) 1 3 1 (Ll Gyraulus corinna 9 34 IB 20 24 67 129 59 21 potamopyrgus antipodarum 212 722 191 25B 209 247 BO 114 129 spbaerium novaezelandiae 6 NEMATODA 4 9 2 5 I 2 Appendix 2.10 Numbera of invertebr~te. per sample (= o.ooa m ) of Elod$4 canadensis collected from the western sampling area (water depth = 2m), September 1976 - October 1977. E2m

Oat. I 2/9176 2/11/76 2/12/76 I 20/1/77 2/3/77 I 6/4/77 10/5/77 I 20/6/77 13/7/77 3/10/77 Dry wt mar.rophyte (9l 1°.770.751.63 0.32 0.42 0.54 0.160.22 0.33, 0.40 0.76 0.33 0.560.460.57; 1.270.660.53 0.77 0.76 0.76,0.59 0.47 0.50 0.25 0.22 0.66 0.14 0.73 0.17

Invertebrate taxon I ~------Chloroh~dra viridi$sima 1 4 13 6 23 294 195 306 13 3& 21 141 190 54 363 62 66 87 240 160 59 1 11 5 14 Cura pinguis 3 2 7 Plu~tell. re~nm I' I' P P P P P P P P P P P P P P P P P P P !' Cnaetogdster 51'. 13 6 3 1 4 29 9 9 1 8 3 3 20 13 16 1 6 19 3 7 12 10 6 5 4 2 other OLlGOCIIAE'l'lI 13 2 3 9 46 41 27 24 14 12 56 3 2 20 21 255 15 5 2 67 J 6 1 Bosmjn~ meridionali$ 6 15 6 61 69 15 6 2 Alana gutt.ata 6 3 1 Craptoleberls te8cudinarja 19 4 92 73 16 30 53 6 10 11 17 5 28 12 2 1 Chydorus sphaericus 1 3 6 2 6 3 5 3 unident. CLADOCERA 3 3 3 1 26 73 2 57 104 23 3 OSTRACODA 1 1 I, Eucyclops serrul4tus 4 ) ) 12 32 3 3 29 5 46 41 27 10 7 4 83 30 40 229 68 107 35 50 21 26 72 J8 26 14 XQnthocnemis 2e41dndic4 1 1 1 2 1 1 1 1 1 " DiiJprepocor.is zedlandiae IA) (Ll 1 1 1 F'dco,x:,tetb..ic.a bendorson.i. (Ll 3 2 1 2 3 1 4 2 1 3 4 5 1 1 1 4 (P) 1 1 P. tilly".d:i (Ll 3 3 precase HYOROPTILIO~ 12 3 9 1 1 2 1 2 1 5 5 1 1 1 Triplectides cephdlotes (L) 1 1 1 CHIRONOMIDAE 1 10 2 2 4 2 4 5 6 52 7 1 3 21 1 3 11 1 4 pjon~ unCd~a exigua (Al 1 8 1 2 2 32 3 2 4 1 4 17 14 12 35 13 13 1 6 2 28 2 (L) 3 1 4 Hydrozetes lemnae 1 3 10 10 5 7 6 3 1 3 2 7 4 3 ~ 2 8 1 1 TriIDaldconOthrus nOVU$ 0\1 11 '8 2 1 2 1 (Ll Arrenurus sp. 1 GyriJulus corinna 10 3 29 4 2 9 12 2 18 27 39 14 15 16 18 90 79 74 50 104 71 6 5 15 14 21 24 3 1 1 potamopyrgus antipodar~ 34 43 44 32 46 91 35 31 65 99 176 110 60 73 67 2SB 221 250 404 299 551 143 154 216 97 60 129 41 136 64 Sp~eri.um nova~zelandiGe 1 1 3 N£MATODA 1 1 2 1 2

.-.~ I - ~~ ------, ---- ,- Appendix :2. 11 Numbers of invertebrates per sample (= 0.008 m2) of Elodea canadensis from the western sampling area (water depth = 3 m), September 1976 - October 1977. I\,) w E o

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Dry wt macrophyte (g) 1.32 0.85 0.40 0.24 0.16 0.46 0.40 C.31 0.19 Invertebrate taxon I !

Chlorohydra viridissima 22 9 22 55 143 2000 89 92 276 Cura pinguis 1 1 6 p Plumatella repens P P P P P P P Cf>.aetogas ter sp. 42 5 5 60 7 other·OLlGOCHAETA 34 2 4 2 3 2 3 Glossiphonia sp. 1 Bosmina meridionalis 15 370 Alona guttata 1 2 1 2 Graptoleberis testudinaria 2 2 3 8 Chydorus sphaericus 1 1 unident. CLAOOCERA 10 8 43 38 OSTRACODA 1 2 Eucyclops serrulatus 10 5 66 21 65 160 60 46 42 xanthocnemis zealandica 2 Sigara arguta (Al 1 Paroxyethira hendersoni (Ll 2 2 1 (P) 2 2 precase HYDROPTILIDAE 10 1 CHIRONOMIDAE (Ll 2 4 40 18 (P) . 1 Piona uncata exigua 7 8 2 15 4 9 Hydrozetes lemnae 34 16 5 6 3 2 Trimalaconothrus novus (Al 2 4 (Ll Gyraulus corinna 4 6 13 6 15 59 65 5 29 potamopyrgus antipodarum 98 121 65 36 71 179 183 116 188 Sphaerium novaezelandiae 1 1 l I I Appendix 2 12 Numbers of invertebrates per sample (= 0.008 m2) of Elodea canadensis collected from the we~tern sampling area (water depth - 4 m), September 1976 - October 1977. E

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 Dry wt macrophyte 3.41 0.52 0.47 0.40 0.53 0.36 0.86 0.28 0.09 Invertebrate taxon

Chlorohydra viridissima 8 1 30 251 440 389 1147 Cura pinguis 53 86 1 4 1 Plumatella repens p p p p p p p p Chaetogaster sp. 5 5 6 9 125 9 other OLlGOCHAETA 9 4 6 8 Bosmina meridionalis 2 2 Alona guttata 10 47 Graptoleberis testudinaria 2 1 Chydorus sphaericus 6 16 2 6 Ceriodaphnia dubia 3 1 unident. CLADOCERA 1 6 4 47 2 OSTRACODA 3 Eucyclops serrulatus 1 1 23 26 102 14 52 Procordulia grayi 125 29 7 , 1 I Paroxyethira hendersoni (L) 2 1 2 1 (P) 1 P. tillyardi (L) 1 precase HYDROPTILIDAE 2 1 CHIRONOMlDAE 1 ..-.;''; 17 Piona uncata exigua 1 3 16 Hydrozetes lemnae 7 4 4 3 2 16 19 15 Trimalaconothrus 3 novus (A) 1 20 13 Gyraulus corinna 7 3 18 41 58 potamopyrgus 108 4 10 48 40 72 41 110 333 435 103 I I 27 Appendix 2.13 Numbers of invertebrates per sample (- 0.008 m2 ) of Isoetes alpinus collected from the western sampling area (water depth'" 0 - 1 m) I September 1976 - October 1977. N W N

Date 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 I 20/6/77 13/7/77 3/10/77\ Dry wt macrophyte (g) 1.40 0.14 0.09 0.23 0.11 0.12 0.18 0.12 0.11

Invertebrate taxon

Chlorohydra viridissima 1 1 3 10 7 16 3 2 Cura pinguis 1 2 14 22 2 3 Plumatella repens P P Chaetogaster sp. 8 1 1 2 28 20 other OLIGOCHAETA 3 1 3 30 7 Bosmina meridionalis 2 4 Alona guttata 18 22 16 Graptoleberis testudinaria 7 16 6 Chydorus sphaericus 1 Ceriodaphnia dubia 5 unident. CLADOCERA 20 1 2 50 71 OSTRACODA 46 12 80 1 1 6 80 27 2 Eucyclops serrulatus 4 46 27 134 30 230 33 Xanthocnemis zealandica 1 1 Oxyetbira albiceps (P) 1 Paroxyethira hendersoni (Ll 11 2 7 1 23 7 16 (P) 115 3 2 P. tillyardi (Ll 3 4 1 precase HYDROPTILIDAE 16 1 5 7 7 6 1 Triplectides cepbalotes (L) 3 2 2 oecetis unicolor (L) 2 1 2 3 1 Oecetis iti (L) 1 Nymphula nitens 3 I CHlRONOMIDAE 2 10 1 4 26 51 43 Arrenurus sp. 1 Piona uncata exigua 1 2 3 1 1. 8 2 Hydrozetes lemnae 32 13 62 12 2 9 4 4 Trimalaconothrus novus (Al 25 19 24 5 1 44 2 (Ll 3 1 9 2 10 9 i 9 1 Gyraulus corinna 5 2 2 10 15 31 83 39 20 potamopyrgus antipodar~~ 440 22 219 98 57 280 317 288 149 Spbaerium novaezelandiae 8 3 26 16 4 NEMATODA 1 1 6 16 25 10 I 1 I I I Appendix 3 Chironomids collected the quantitative sampling program (see Appendix 2 for numbers of samples collected).

2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 TOTAL

NE 2m Ablabesmyia .mala 1 1 Macropelopia/Gressittius spp. 4 1 2 1 1 9 Cricotopus spp. 7 1 3 10 6 34 10 71 Chironomus zealandicus 1 1

SI 1m Ablabesmyia mala 8 1 9 Macropelopia/Gressittius spp. 4 3 3 2 12 Cricotopus spp. 2 3 7/1 pupa 6 27 23 1 69/1 pupa Orthocladiinae A 4 5 3 12 Orthocladiinae B 1 3 4 Orthocladiinae C 1 1 Chironomus zealandicus 1 1 Tanytarsus vespertinus 1 15 8 4 3 31

Cont'd

tv W W 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 TOTAL

SM 1m mala 1 1 pupa 1/1 pupa ttius spp. 1 1 3 1 6 spp. 2 21/1 pupa 6 47 33 4 113/1 pupa Orthoc1adiinae A 1 4 3 4 3 15

Orthoc1adiinae B 1 4 5 Orthoc1adiinae C 2 1 3 vespertinus 10 1 3 25 14 53

EEl m Macropelopia/Gressittius spp. 2 6 8 spp. 2 2 Chironomus zealandicus 1 1

EE 2m mala 2 2

Macropelopia/Gressittius spp. 4 2 7 5 6 2 9 11 1 47 Cri spp. 3 1 4 8

Cont'd 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 TOTAL

WE 2m Macropelopia/Gressittius spp. 9 2 4 1 3 5 24 Cricotopus spp. 1 4 6 13 56 2 22 11 115 Chironomus zealandicus 1 1 2

WE 3m Macropelopia/Gressittius spp. 2 2 Cricotopus spp. 2 4 40 16 62 Syncricotopus pluriserialis 1 pupa 1 pupa

WE 4m Cricotopus spp. 1 34 16 51 Chironomus zealandicus 1 1

WI 1m Macropelopia/Gressittius spp. 12 2 11 4 29 Cricotopus spp. 1 4 20 33 37 95 Tanytarsus vespertinus 4 7 2 13

N W 111 tV W 0\ 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77 TOTAL

EI 1m pentaneura sp. 2 2 mala 2 2 1 2 2 9 Macropelopia/Gressittius spp. 2 2 Cricotopus spp. 14 6 17 8 45

EM mala 2 2 Cricotopus spp. 4 10 2 6 1 23

ER Cricotopus spp. 1 1 9 11 Orthoc1adiinae B 1 1

WE 1m Macropelopia/Gressittius spp. 2 2 1 1 1 1 8 spp. 3 75 6 10 94 Tanytarsus vespertinus 1 1

Cont'd Appendix 4.1 Indiesi of precision (0) for macrophyte dry weight and invertebrate density obtained from triplicate (n .. 3) samples from N E 2 during the quantitative s~pling program 12 September 1976 - 3 October 1977, Appendix 2.1) and the pilot sarvey (14 April 1976, Appendix 1) with densities expressed as numbero/sample (numbers/g dry wt: of INlcrophyte) •.

Pilot 14/4/76 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77

Ory wt 0.21 0.51 0.19 0.12 0.10 0.20 0.07 0.15 0.18 0.29 0.11 COEL.E:NTERA'l'A 0.87(0.72) 0.66(0.50) 0.88(0.85) 0.75(0.67) 0.09(0.16) 0.29(0.15) 0.23(0.25) 0.51(0.48) 0.32(0.17) 1.00(1.00) ANNELIDA 0.56(0.47) 0.95(0.56) 1.00(1.00) 0.52(0.49) 0.31 (0. 23) 0.26(0.32) 0.36(0.32) 0.60(0.88) 0.37(0.28) 0.21(0.82) 0.35 (0. 32) CIWSTACEA 0.42 (0. 53) 0.55(0.92) 0.38(0.34) 0.38 (0. 39) 0.15 (0.17) 0.27(0.35) 0.06(0.003) 0.22(0.45) 0.34(0.36) 0.24 (0.48) 0.39 (0. 40) INSECTA 0.67(0.66) 0.29 (0. 55) 0.74(0.71) 0.33(0.22) 1.00(1.00) 0.50(0.51) 0.26(0.33) 0.24(0.31) 0.06(0.16) 0.56(0.40) 0.16(0.16) ACARINA 0.39(0.27) 1.00 (1. 00) 0.19(0.32) 0.36(0.36) 0.28(0.23) 0.06(0.47) 0.22(0.19) 0.18(0.77) 0.23 (0. 28) 1.00(1.00) 0.38(0.39) MOLLUSCA 0.25(0.21). 0.40(0.21) 0.14(0.09) 0.17(0.11) 0.19(0.11) 0.37(0.20) 0.07(0.07) 0.31(0.11) 0.14(0.12) 0.34(0.32) 0.09(0.17) Total Invertebrates 0.25(0.22) 0.48(0.17) 0.16(0.10) 0.21(0.13) 0.13(0.08) 0.16(0.04) 0.05(0.03) 0.20(0.32) 0.17(0.15) 0.25(0.41) 0.19(0.06)

Appendix 4.2 Indices of precision (0) for INlcrophyte dry weight and invertebrate density obtained from duplicate (n - 2) samples from S I during the qaantitative sampling program (2 September 1976 - 3 October 1977, Appendix 2.2) and the pilot sarvey (14 April 1976. Appendix 1) with densities expressed as numbers/sample (numbers/g wt of macrophyte) •

pilot 14/4/76 2/11/76 2/3/77 8/4/77 13/7/77

Dry wt: 0.35 0.16 0.41 0.06 0.60 COELENTERATA 0.75(0.87) 0(0.16) . 0.18(0.64) 0.33 (0. 38) 1. 00 (1. 00) ANNELIDA 0.88(0.64) 0(0.41) 0.22(0.17) 0.50(0.85) CRUSTACEA 0.71 (0. 85) 0(0.16) 0.28(0.14) 0.43(0.39) 0.31(0.77) INSECTA 0.35(0.C04) 0.10(0.07) 0.43(0.71) 0.09(0.03) 0.33(0.33) ACA.RlNA 0.10(0.43) 0.02 (0.15) 0.10(0.49) 0(0.06) 1.00 (1.00) MOLLUSCA 0.59(0.30) 0.06(0.10) 0.01(0.42) 0.26(0.31) 0.33(0.33) Total Invertebrates 0.53(0.21) 0.23 (0.07) 0.03(0.43) 0.15(0.21) 0.006(0.60) tv w --..I Appendix 4.3 Indices of precision (Ol for ~crophyte dry weight and invertebrate density obtained from triplicate (n " 3) samples from E E 2m during 1'>J the quantitative sampling program (2 September 1976 • 3 October 1977, Appendix 2.5) and the pilot survey (14 april 1976, Appenjix 1) w with densities expressed as numbers/sample (numbers/g dry we of macrophyte). CD

Pilot 14/4/76 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 3/10/77 cry we 0.08 0.01 0.03 0.19 0.15 0.24 0.16 0.08 0.27 0.20 COELENTERATA 0.45(0.42) 1.00 (1.00) 0.76(0.75) 0.15(0.23) 0.24(0.42) 0.51(0.28) 0.29(0.16) 0.46(0.52) 0.70(0.52) 0.64(0.60) ANNELIDA 0.81(1.01) 0.97(0.97) 0.84(0.84) 0.43 (0, 50) 0.50(0.75) 0.29(0.24) 0.66(0.67) 0.50(0.46) 0.35(0.65) 0.33 (0.45) CRUSTACEA 0.43(0.54) 0.18(0.19) 0.61(0.61) 0.35 (0. 58) 0.29 (0.22) 0.21(0.03) 0.36 (0. 42) 0.47(0.44) 0.29(0.46) 0.26(0.42) INSECTA 0.18(0.12) 0.13(0.34) 0.13(0.15) 0.14 (0.50) 0.44(0.69) 0.42(0.36) 0.21(0.31) 0.42(0.41) 0.26(0.35) 0.24(0.30) ACAIUNA 0.19(0.43) 0.45(0.44) 0.17(0.19) 0.13(0.30) 0.38(0.45) 0.30(0.11) 0.25(0.33) 0.64(0.60) 0.17(0.52) 0.23(0.26) MOLLUSCA 0.26(0.29) 0.25 (0.25) 0.22(0.24) 0.12(0.27) 0.09(0.25) 0.36{0.25) 0.21(0.26) 0.11(0.13) 0.17{0.17) 0.17(0.33) Total Invertebrates 0.24(0.26) 0.23(0.23) 0.21 (0.23) 0.11(0.25) 0.08(0.26) 0.35(0.22) 0.21(0.25) 0.16(0.16) 0.12 (0.23) 0.10(0.27)

Appendix 4.4 Indices of precision (0) for macrophyte dry weight and invertebrate density obtained from duplicate (n - 2) samples from EI during the quantitative sampling program (2.September 1976 • 3 October 1977, Appendix 2.6) and the pilot survey (14 April 1976, Appendix 1) with densities expressed as numbers/sample (numbers/g dry wt of macrophyte).

pilot 14/4/76 219/76 2/11/76 2/12/76 2/3/77 20/6/77

Dry wt 0.22 0.09 0.14 0.40 0.33 0.25 COELENTERATA 0.90 (0.63) 1.00(LOO) 0.52(0.22) 1.00 (1. 00) ANNELIDA 1.00 (1. 00) 1. 00 (1. 00) LOO (1.00) 0(0.25) CRUSTACEA 0.72(0.86) 0.80(0.77) 0.52(0.62) 0.34(0.07) 0.90(0.81) 0.33(0.53) INSECTA 0.58(0.70) 0.83(0.81) 1.00(LOO) 0.51 (0.14) 0.90(0.81) 0.50(0.29) ACAIUNA 0.77(0.91) 0.90(0.69) 0.73 (0.79) 0.23 (0.19) 0.71 (0.50) LOO{l.OO) MOLLUSCA 0.48(0.60) 0.80(0.77) 0.44(0.54) 0.85(0.67) 0.71(0.49) 0.02(0.27) Total Invertebrates 0.51(0.64) 0.83(0.80) 0.54 (0.63) 0.49(0.11) 0.76(0.58) 0.15(0.39) Appendix 4,S Indicee of precision (D) for macrophyte dry weight and invertebrate density obtained from replicate In .. 3) samples from WE 2 III during the qU4ntitstive sampling program !2 September 1976 - 3 October 1977. Appendix 2.10) and the pilot survey (14 April 1976, Appendix 11 with densities expressed as numbers/sample (numbers/g dry 10ft of macrophytel •

Pilot 14/4/76 2/9/76 2/11/76 2/12/76 20/1/77 2/3/77 8/4/77 10/5/77 20/6/77 13/7/77 3/10/77

Dry wt 0.15 0.28 0.15 0.18 0.28 0.06 0.15 0.004 0.07 0.38 0.55 COEl.EN'l'EAATA 0.91(0.75) 0.12(0.57) 0.35(0.28) 0.13(0.32) 0.29(0.35) 0.31(0.26) 0.62(0.62) 0.26(0.31) 0.98(0.97) 0.27(0.44) ANNELIDA 1.00(1.05) 0.1!:(O.31) 0.48(0.38) 1.00(1.00) 0.21(0.30) 0.21(0.26) 0.42(0.19) 0.12 (0. 72) 0.13 (0.13) 0.63(0.29) 0.30(0.47) CRUSTACEA 0.6010.57) 0.05(0.24) 0.51(0.48) 0.62 (0.61) 0.27(0.15) 0.36(0.42) 0.30(0.07) 0.26(0.26) 0.17(0.21) 0.31(0.49) 0.38(0.77) INSECTA 0.59(0.60) 0.29(0.36) 1.00(1.00) 0.58(0.54) 0.40(0.47) 0.05(0.05) 0.11(0.36) 0.62(0.61) 0.33(0.36) 0.57 (0. 75) 0.60{0.70) ACARINA 0.35(0.45) 0.66(0.52) 0.30(0.32) 0.18(0.33) 0.64(0.64) 0.13(0.18) 0.11(0.18) 0.26(0.26) 0.58(0.59) 0.50(0.67) 0.86{0.51)

MOLLUSCA 0.38(0.33) 0.17(0.09) 0.32(0.17) 0.27(0.16) 0.19(0.09) 0.14(0.15) 0.04(0.25) 0.13(0.14) 0.14(0.17) 0.18(0.18) 0.34 (0.19)

Total Invertebrates 0.26(0.36) 0.13 (0.13) 0~18(0.03) 0.15 (0.20) 0.06(0.21) 0.12(0.18) 0.13(0.12) 0.18(0.17) 0.13(0.19) 0.04(0.26) 0.22(0.28)

Appendix 4.6 Indices of precision (D) for macrophyte dry weight and invertebrate density obtained from duplicate (n = 2) samples from EM, E Rand WI from the quantitative sampling program (2 September 1976 - 3 October 1977, Appendices 2.7, 2.8 and 2.13) and the pilot survey (14 April 1976, Appendix 1) with densities expressed as numbers/sample (numbers/; dry wt of macrophyte).

EM EM ER ER W.I, 2/11/76 2/12/76 14/4/76 2/12/76 2/11/76

Dry 'lit 0.13 0.19 0.22 0.04 0.22 COEl.EN'l'EAATA 1.00(1.00) 0.38(0.52) 1.00(1.00) 0(0.22) ANNELIDA 1.00(1.00) 1.05 (1.17) 0(0.04) 0.45(0.26)

CRUSTACEA 1.00(1.00) 0.02(0.20) 0.40(0.34) 0(0.04) 0.78(0.85) INSECTA 0.33(0.44) 0.15(0.33) 0.22(0.05) 0.22(0.26) 0.43(0.59) ACARINA 0(0.13) 0.03(0.22) 0.37(0.20) 0.11(0.15) 0.71(0.80) MOLLUSCA' 0.12(0.007) 0.34 (0.17) 0.33(0.11) 0.25(0.22) 0.Bl(0.87) Total Invertebrates 0.05(0.08) 0.22(0.04) 0.32(0.10) 0.15(0.11) 0.73(0.82) rv w \..0 N ,g:,. o

Appendix 5.1 Adult Chironomidae collected by hand-net sweeps adjacent to Lake Grasmere, 2 November 1976 - 7 December 1978. All collections from the southern end of the lake exc~pt 14-20/1/77 (north and east), 27/10/78 (north and south) , 17/11/78 and 7/12/78 (north) . (males., females)

1976 1977 1978 2/11 2/12 14-20/1 2/3 8/4 10/5 3/10 29/11 26/12 11/4 9/10 27/10 17/11 7/12

TANYPODINAE Ablabesmyia mala 0,1 0,2 0,1 Gressittius antarcticus 7,1 0,1 1,2 0,1 Macropelopia uinbrosa 2,1 2,7 . 3,0 2,14

PODONOMINAE Parochlus sp. 0,1 0,2 2,13 0,3

ORTHOCLADIINAE Syncricotopus 0,1 0,2 pluriserialis 0,4 1,0 0,1 1,1 1,,3 0,11 30,18 9,5 2,3 Cricotopus zealandicus 0,2 Metriocnemis lobifer 3,0 1 a Eukiefferiella sp. 1,0 2,7 70,31 24,10 "black species" 1,0 9,29

CHI RONOMINAE Chironomus zealandicus 5 5 3,2 4,4 1,1 1,1 3,1 9,7 3,2 9,2 23,7 8,1 19,23 xenochironomus 2,0 canterburyensis 5,0 1,0 1,0 1,0 5,3 Cladopelma curti valva 3,5 2,1 Polypedilum canum 10,0 1,2 Tanytarsus vespertinus 5,0 1,0 2,0 114,6 1,0 Calopsectra funebris 1,0 . Appendix 5.2 Adult Chironomidae collected by light trapping at the southern end of Lake Grasmere. (males ,females) * = generator-powered mercury discharge lamp (160W) ** = battery-powered (12V) fluorescent tube (white light) All times given are N.Z.D.T.

29/11/77 26/12/77 25/1/78 1. 7/12/78 1h between 2045- 2245- 2345- 2200-2300h 2300-2400h 2300 & 0300h 2245h 2345h 0045h * * * ** * ** '* *'* *

TANYPODINAE Ablabesmyia mala 1,2 0,1 0,1 Gressittius antarcticus 0,1 0,235 0,13 1,24 0,5 0,7 0,15 0,47 Macropelopia languidus 0,1 0,2 M. umbrosa 0,11 0,5 0,2 0,1 0,41

PODONOMINAE Parochlus sp. 0,1 1,1

ORTHOCLADIINAE Syncricotopus pluriserialis 0,2 0,3 0,99 0,51 0,15 0,10 0,15 0,46 0,11 Cricotopus zealandicus 0,1

CHIRONOMINAE

Chironomus zealandicus 5,3 1,2 7,383 0,11 0,1 5,1 2,4 0,2 18,40 xenochironomus canterburyensis 2,7 0,1 10,2 Cladopelma curti valva 0,1 Polypedilum pavidus 0,1 1,12 Tanytarsus vespertinus 1,0 0,1 2,0 Appendix 5.3 Adult Trichoptera collected by hand-net sweeps adjacent to Lake Grasmere, 2 November 1976 - 7 December 1978. Collection site details given in Appendix 5.1 (males ,females) .

1976 1977 1978 2/11 2/12 14-20/1 2/3 8/4 10/5 3/10 29/11 26/12 11/4 9/10 27/10 17/11 7/12

HYDROPSYCHIDAE Aoteapsyche colonica 2,2 1,0 A. tepoka 1,0

CONOESUCIDAE Pycnocentrodes aureola 8,4 0,1 8,1 9,0 0,1 1,0 15,0

HYDROPTILIDAE Oxyethira albiceps 0,1 0,1 0,2 0,1 0,1 2,1 5,1 4,1 paroxyethira hendersoni 1 / 0 17,27 5,7 65,9 10,3 6,2 37,4 90,18 8,3 9,18 53,S 83,4 9,3 o 2

P. ti11 yardi 44,24 0 1 2 0,1 3,1

LEPTOCERIDAE Oecetis unicolor 10,12 Hudsonema amabilis 2,15 6,33 0,1 1,4 4,0 10,1 Triplectides cephalotes 0,1 T. obsoleta 0,1 1,3 1,0 Appendix 5.4 Adult Trichoptera collected by light trapping at the southern end of Lake Grasmere. (males, females) * = generator-powered mercury discharge lamp (160W) ** = battery-powered (12V) fluorescent tube (white light) All times given are N.Z.D.T.

29/11/77 26/12/77 25/1/78 7/12/78 Ih between 2045- 2245- 2345- 2200-2300h 2300-2400h & 0300h 2245h 234Sh 004Sh 2300 * * * **" * ** * ** 1<

HYDROPSYCHIDAE Aoteapsyche colonica 1,0 2,0 2,0 1,0 A. tepoka 6 1 13,0 3,0 7,0 9,1 4 0 POLYCENTROPODIDAE polyplectropus puerilis 2,0 1,0 18,9 RHYACOPHILIDAE Hydrobiosis umbripennis 1,2 10,17 3,6 1,6 1,0 H. parumbripennis 0,6 H. harpidiosa 4,0 1,0 H. frater 2,0 11,0 3,0 5,1 0,1 H. clavigera 3,0 1,0 1,0 psilochorema bidens 1,0 2,0 Ps. leptoharpax 2,1 1,0 17,0 1,0 1,0 0,2 Costachorema xanthoptera 0,2 4,0 15,17 2,0 1,0 1,1 1,0 C. callis tum 1,0 CONOESUCIDAE pycnocentrodes aureola 1,3 2,3 5,2 8,10 4,8 111,329 HYDROPTILIDAE Oxyethira albiceps 0,3 0,1 21,45 4,22 30,62 19,24 150,183 0,11 38,36 Paroxyethira hendersoni 21,2 17 ;2 93,160s 20,13 134,40 6,37 63,28 4,22 24,16 P. tillyardi 1,1 LEPTOCERIDAE Oecetis unicolor 4,0 5,0 5,0 16,3 21,2 Hudsonema amabilis 1,1 2,0 1,0 4,2 I\.) 3,1 15,12 4,8 202,57 .t:> Triplectides cephalotes 4,0 1,0 2,0 w s = sub sample Appendix 5. 5 Adult Tipulidae collected by hand-net sweeps adjacent to Lake Grasmere, 2 December 1976 - 2.February.1979. (Collection site details given in Appendix 5.1.)

2/12/76 20/1/77 2/3/77 3/10/77 26/12/77 ·11/4/78 9/10/78 27/10/78 17/11/78 7/12/78 2/2/79

Zelandotipula sp. 1 1

Leptotarsus (Macromastix) 1 minutissima

Limonia (Dicranomyia) 1 otagensis

L. (D.) sp. A 1

L. (D.) sp. C 1

?Metalimnophila 1

Apbrophila neozelandica 1

Erioptera (Trimicra) 20+ 1 1 1 1 9 1 1 pilipes

Amphineurus 4 spp. 1 1 1 1 1 1

Molophilus sp. 1 1 1 1 Appendix 5.6 Adult Tipulidae collected by light trapping at the southern end of Lake Grasmere. * = generator-powered mercury discharge lamp (160W) ** = battery-powered (12V) fluorescent tube (white light) All times given are N.Z.D.T.

26/12/77 25/1/78 7/12/78 2200-2300h 2300-2400h Ih between 2300 & 0300h *' * ** "*

Limonia (Dicranomyia) ? vicarians 1

Aphrophila neozelandica 1 1

Erioptera pilipes 1 1 1

Amphineurus 2 sp. 2 1 1 Appendix 5.7 Adults of Odonata, P1ecoptera and Ephemeroptera collected ~y hand-net sweeps adjacent to Lake Grasmere, 2 November 1976 - 7 December 1978. (Collection site details given in Appendix 5.1) (ma1es,fema1es)

1976 1977 1978 2/11 2/12 14-20/1 2/3 8/4 3/10 29/11 9/10 27/10 17/11 7/12

ODONATA xanthocnemis zealandica 2,0 16,13 7,6 1,0 5,4

PLECOPTERA

Zelandobius furcillatus 3 1 0 I,D 1,0 4,1 5,4 7,0 4 0

EPHEMEROPTERA Deleatidium sp. 0,2 0,6 0,1 44,1

1~ Deleatidium light trapped 26/12/77, 2200-2300h NZDT in mercury vapour trap. 247

6.1 (a) Percentage composition of the faecal material of final instar Paroxyethira hendersoni,in terms of major food categories, on a projected area basis; and (b) generic composition (% by numbers) of the diatom category. *:=<0.1%. (September 1976 - November 1977 and overall)

1976 1977 Over- 2/9 2/11 2/12 20/1 8/4 20/6 18/8 21/11 all

Diatoms 54.8 37.5 53.1 49.9 41.0 48.6 67.5 52.0 50.0 Detritus 19.5 30.6 14.3 22.5 18.0 11.3 14.5 24.4 18.9 Macrophyte 7.9 5.9 4.4 9.4 4.7 14.8 2.4 13.5 7.2 Filamentous algae 17.8 26.0 28.2 17.3 26.2 16.1 15.6 10.1 21.3 S.R.T's 0.9 10.1 9.2 2.7

Number of larvae 12 11 10 10 10 9 7 4 72 examined

(b)

Cocconeis 34.7 57.6 73.8 59.1 93.9 82.3 68.9 28.1 63.0 Gomphonema 27.9 13.6 4.4 0.9 1.6 0.3 14.1 62.9 15.8 Epithemia 6.6 15.7 17.0 34.8 1.1 5.4 1.5 0.4 7.9 Synedra 11.2 0.3 0.3 1.3 5.0 2.7 3.5 Asterionella 6.2 3.8 6.2 2.6 Fragilaria 2.7 5.8 2.1 3.8 1.5 5.8 2.6 Rhoicosphenia 4.2 7.3 4.4 4.3 0.3 2.3 Cymbella 5.2 0.3 0.3 1.1 Cyclotella 1.2 0.9 0.8 1.4 2.6 0.7 Melosira 0.7 2.1 0.7 Navicula 0.7 * 248

Appendix 6.2 (a) Percentage composition of the faecal material of final instar Paroxyethira tillyardi, in terms of major food categories, on a projected area basis; and (b) generic composition (% by numbers) of the diatom category. (September 1976 - November 1977 and overall)

1976 1977 2/9 2/11 2/12 8/4 20/6 18/8 21/11 Overall

(a) Diatoms 58.1 40.5 57.0 31.7 23.5 45.0 31.3 41.0 Detritus 17.2 33.5 2l.5 21.1 23.5 33.9 33.9 26.4 Macrophyte 4.5 2.2 9.1 11.9 13.1 7.3 7.5 8.0 Filamentous algae 20.2 23.8 12.4 25.2 39.2 13.8 27.3 23.1 S.R.T's 10.1 0.7 l.5

Number of larvae 12 10 10 9 10 7 16 74 examined

(b) Cocconeis 75.3 6l.6 74.6 82.0 95.5 91.3 76.4 77 .8 themia 5.8 2l.2 8.5 4.6 0.9 1.0 3.6 7.3 Rhoicosphenia 3.5 3.4 12.3 9.3 1.8 3.8 6.2 Gomphonema 13.1 10.3 2.9 0.7 2.9 12.7 6.1 Fragilaria 0.7 2.7 0.9 l.0 5.5 1.0 Synedra l.2 2.7 0.7 Cyclotella 0.4 0.7 0.9 0.4 Navicula 0.8 0.7 0.3 Cymbella 0.7 0.1 Pinnularia 0.4 0.1 Appendix S.3 Percentage composition of the faecal material of Hudsonema amabilis, in terms of major food categories, on a projected area basis (2 September 1976 - 21 November 1977 and overall) for ins tars 2 - 5.

2nd ins tar 3rd ins tar 8 Apr. 20 Jun. 18 Aug. OVERALL 2 Sep. 2 Nov. 8 Apr. 20 Jun. 18 Aug. OVERALL

Diatolns 27.2 26.2 33.9 28.8 10.8 42.3 40.1 23.6 27.3 36.3 Detritus 50.6 40.3 52.3 46.3 32.9 27.5 34.9 20.9 28.3 25.9 l'IzIcrcphyte 9.2 4.0 2.0 4.4 28.8 13.8 13.2 19.6 3.4 14.1 Filamentous algae 12.3 10.1 11.8 11.1 11.0 16.5 11.0 23.4 11.9 16.6 Animal 0.4 0.2 16.5 0.7 2.5 o.a 2.6 5.:I<.T's 0.7 19.0 9.1 9.9 28.3 4.5

Number of larvae 2 3 6 11 3 1 2 7 examined 5 la

4th ins tar 5th ins tar 2 Sep. 2 NOV. 20 Jan. 8 Apr. 20 Jun. 18 Aug. 21 Nov. OVERALL 2 Sep. 2 Nov. 20 Jan. 8 Apr. 20 Jun. 1a Aug. 21 Nov. OVERALL

DiatolllS 56.2 30.6 35.5 39.0 30.9 48.4 32.2 42.4 42.5 32.3 32.8 22.4 32.6 38.7 24.9 33.5 Detritus 17.8 25.7 21.3 19.8 24.2 21.8 23.3 20.9 17.4 18.4 19.8 13.5 24.0 22.2 20.8 16.8 l'IzIcrophyte 20.2 21. 2 14.9 20.9 12.2 12.0 13.5 16.6 34.5 29.5 21.3 55.2 19.9 22.1 19.1 29.0 Filamentous algae 4.3 9.1 4.3 5.7 15.9 12.7 15.6 9.4 6.1 8.3 6.4 5.0 10.3 3.1 14.3 8.2 Animal 1.5 13.4 21.3 19.4 10.3 4.8 12.7 9.2 9.5 10.4 19.5 4.0 11.4 a.s 19.6 12.4 S.R.T's 2.7 6.5 0.3 2.8 1.5 0.2 1.9 0.1 1.4 0.4

Number of larvae 10 4, 1 10 10 10 5 50 10 10 10 6 10 10 10 66 examined

IV '""\0 Appendix 6.4 Generic composition (\ by numbers) of diatoms in the faeces of Hudsonema amabilis (2 September 1976 - 21 November 1977) w < o.a. N (See Appendix 6.3 for number of larvae examined.) oV1

2nd instar 3rd ins tar

8 Apr. 20 Jun. 18 Aug. OVERALL 2 Sep. 2 Nov. 8 Apr. 20 Jun. 18 Aug. OVERALL

Cocconeis 76.9 65.4 95.5 81. 2 66.7 36.6 84.8 60.5 63.9 63.6 Asterionella 3.8 1.4 6.6 2.4 12.8 16.5 9.3 Rhoicosphenia 1l.5 9.6 5.6 8.6 12.2 2.2 9.3 4.1 7.2 r::pithemia 7.7 7.7 4.2 6.8 24.4 10.9 7.0 6.2 9.0 cyclotella 3.8 1.4 2.2 1.2 0.5 Gomphonema 3.8 1.5 2.1 6.8 2.1 2.6 Fragilaria 3.9 1.9 1.4 0.9 17.1 2.3 3.1 3.4 Navicllla 0.9 4.9 3.5 2.1 2.1 Synedra 1.9 0.7 1.7 2.4 2.1 1.3 Cymbella 1.9 0.7 2.3 0.5 Diatoma 0.9 0.3 Rhopalodia 3.0 1.4 Pinnularia 1.2 0.3

4th ins tar 5th ins tar

2 Sep. 2 Nov. 20 Jan. 8 Apr. 20 Jun. 18 Aug. 21 Nov. OVERALL 2 Sep. 2 Nov. 20 Jan. 8 Apr. 20 Jun. 18 Aug. 21 Nov. OVERALL

Cocconeis 66.5 47.8 65.2 67.8 43.0 59.7 84.0 62.9 69.4 69.9 49.6 85.2 68.5 53.3 77.9 65.8 Asterionella 5.0 32.3 14.5 9.2 9.2 0.7 0.8 1.7 11.1 18.3 6.2 Rhoicosphenia 3.3 18.5 21. 7 5.3 13.3 6.6 2.7 6.5 7.1 15.4 9.8 2.9 2.7 2.4 2.3 7.0 Epithemia 5.8 13.0 8.7 17.4 2.5 4.5 4.8 7.1 5.0 8.9 3.9 5.3 4.4 4.2 5.8 5.3 Cyclotella 0.8 1.6 0.4 0.5 3.2 1.2 0.2 29.7 4.1 1.1 0.3 6.3 Gomphonema 9.8 9.8 3.9 1.8 4.0 1.6 5.8 3.9 2.8 2.3 0.8 1.6 2.1 2.9 2.7 Fragilaria 4.0 2.2 1.4 5.9 1.1 3.1 2.3 0.4 2.3 0.8 3.8 7.2 1.2 2.5 Navicula 3.0 5.4 1.3 0.4 1.4 0.5 1.9 0.8 0.9 1.5 1.6 1.6 4.1 1.4 Synedra 1.7 3.3 0.3 1.1 2.1 1.4 0.9 0.3 0.4 1.6 6.1 0.6 1.2 Cymbella 0.8 1.6 0.4 0.7 0.5 0.8 0.4 0.2 0.4 1.6 0.3 0.4 Ellnot:ia 0.7 1.6 0.2 0.7 1.6 3.2 0.6 Diatoma 0.7 ." 1.0 2.5 0.5 Melosira 0.3 1.5 0.2 Rhopalodia 0.3 '" Pinnlllaria 0.3 0.3 " Appendix 6.5 Percentage compcsition of the faecal material of Triplectides cephalotes, in terms of major food categories. on a projected area basis (2 September 1976 - 21 November 1977 and overall) for instars 2 - 5.

2nd ins tar 3rd ins tar 4th instar 8 Apr. = OVERALL 2 Nov. 8 Apr. 18 Aug. OVERALL 2 Nov. 2 Dec. 20 Jan. 20 Jun. 18 Aug. 21 Nov. OVERALL

Diatoms 17.5 22.3 9.7 21. 5 20.0 28.0 35.5 21.6 24.9 18.8 25.3 29.1 Detritus 24.0 32.5 23.7 17.1 19.7 12.0 11.8 4.6 21.0 16.1 12.0 1:?1 Macrophyte 38.1 25.4 23.7 38.9 35.5 54.1 36.4 68.3 40.3 47.2 51. 2 47.4 Filamentous algae 10.0 23.7 11.4 11.8 5.9 3.6 4.6 12.4 17.2 11. 5 6.4 Animal 5.0 19.8 11. 8 4.9 7.4 7.7 2.6 S.R.T's 5.4 7.5 6.1 5.6 6.8 0.9 1.4 0.8 2.4

Number of larvae 2 1 1 6 8 8 8 1 2 2 2 23 examined

5th ins tar

2 Sep. 2 Nov. 2 Dec. 20 Jan. 8 Apr. 20 Jun. 18 Aug. 21 Nov. OVERALL

Diatoms 62.3 16.3 22.2 22.8 15.1 17.5 41. 0 8.1 22.1 Detritus 12.7 5.7 12.9 7.4 11.1 11.2 14.6 11.2 9.6 Macrophyte 21. 7 71.9 58.9 55.4 43.7 59.8 31. 7 76.1 57.9 Filamentous algae 2.3 1.4 5.0 5.4 26.6 4.8 11.2 5.3 Animal 1.0 4.6 0.3 9.0 1.2 4.1 4.1 S.R.T's 0.7 0.1 3.6 5.6 1.5 0.5 1.0

Number of larvae 31 examined 4 5 4 6 3 4 2 3

N V1 i-' Generic composition (% by numbers) of diatoms in the faeces of Triplectides cephalotes (2 September 1976 ~977) < N Appendix 6.6 21 November • 0.1%. VI (See Appendix 6.5 for number of larvae examined.) N

2nd ins tar 3rd ins tar 4th instar 8 Apr. = OVERALL 2 Nov. 8 Apr. 18 Aug. OVERALL 2 Nov. 2' Dec. 20 Jan. 20 Jun. 18 Aug. 21 Nov. OVERALL

Cocconeis 76.7 82.6 81. 2 34.5 43.2 74.4 80.5 80.0 79.6 21. 7 45.2 74.4 Gomphonema. 4.4 6.3 17.8 15.5 9.4 5.1 3.7 30.0 7.6 Rhoicosphenia 3.3 6.3 5.8 5.2 4.9 3.0 1.8 1.9 3.3 3.6 Epithemia 6.7 13.0 3.4 4.2 4.3 5.5 5.5 7.4 5.0 48.4 6.1 Fragilaria 6.3 26.4 22.1 1.5 2.6 3.6 1.9 13.3 :<.G Asterionella 3.3 2.9 2.3 2.3 0.2 1.9 5.0 1.3 Cyclotella 0.4 7.3 1.9 5.0 6.5 0.9 synedra 6.7 5.8 4.7 0.8 1.9 0.4 Cymbella 3.3 1.1 0.9 2.1 0.4 3.3 1.2 Navicula 1.1 0.9 0.2 1.8 13.3 0.8 Melosira 2.3 1.0 Opephora 0.2 Pinnularia 0.6 0.5 0.2 " Achnanthes 0.6 0.5

5th ins tar

2 Sep. 2 Nov. 2 Dec. 20 Jan. 8 Apr. 20 Jun. 18 Aug. 21 Nov. OVERALL Cocconeis 27.0 84.3 79.5 74.9 72.8 57.9 36.8 74.2 63.2 Gomphonema. 20.4 1.8 1.8 1.5 2.5 2.9 27.1 3.1 7.5 Rhoicosphenia 31.8 6.4 0.9 3.7 5.0 4.2 2.1 8.1 Epithemia 5.1 2.2 2.7 9.4 6.2 5.7 2.8 7.2 5.5 Fragilaria 2.7 0.7 2.3 8.6 5.7 6.2 1.0 2.3 Asterionella 6.3 9.9 5.0 1.1 18.6 7.2 5.5 Cyclotella 0.9 12.2 2.9 2.1 5.2 4.1 Synedra 0.9 0.7 3.7 0.7 13.9 1.7 Cymbella 3.0 0.4 1.4 4.2 1.1 Navicula 0.9 0.5 1.2 2.8 0.5 Opephora 0.3 0.5 1.2 0.7 0.2 Diatoma. 0.3 " Tabellaria 0.3 " Appendix 6.7 (a) Percentage composition of the faecal material of Nymphula nitens, in terms of major food categories, on a projected area basis (25 July 1976 - 1B August 1977 and overall), for each of three size classes; and (b) Generic composition (% by numbers) of the diatom category.

Siz.e group 1 Size group 2 Size group 3 25 Jul. 2 Sep. 2 Nov. 1B Aug. OVERALL 25 Jul. 2 Sep. 2 Nov. B Apr. 20 Jun. 1B Aug. OVERALL 2 Nov. 20 Jan. 20 Jun. 1B Auo. OVERALL

(al Diatoms 2B.7 6.9 6.B 7.3 11.4 14.1 16.3 10.2 14.5 13.9 12.6 13.5 9.9 17.6 20.5 3.B 14.4 Detritus 36.B 53.8 22.2 30.0 40.7 15.5 20.5 22.3 5.3 35.7 22.4 21.4 1B.6 3.2 31.6 13.0 10.7 Macrophyte 32.9 23.5 66.6 51.8 38.0 69.4 51. 5 60.0 75.7 45.5 60.5 57.4 65.0 76.5 47.9 75.4 71.0 Filamentous algae 1.6 15.9 4.4 10.9 9.9 0.9 11.7 7.5 4.5 5.0 4.5 7.7 6.5 2.7 7.9 3.7 Number of larvae 3 3 1 1 8 2 6 9 2 6 3 28 3 2 1 1 7 examined

(b)

Cocconeis 68.8 62.5 55.6 100.0 68.4 92.4 41.4 82.1 89.3 85.4 77.8 68.2 96.6 99.1 90.5 54.5 96.5 Diatoma 25.4 1.2 10.2 Gomphonema 16.9 10.0 12.5 9.0 1.7 2.4 0.8 4.4 1.7 0.3 0.4 Epithemia 2.6 5.0 11.1 3.7 3.8 2.1 8.5 3.6 4.1 14.8 4.5 1.7 0.3 2.7 27.3 1.5 Asterionella 14.4 0.9 1.6 6.1 2.7 0.4 Fragilaria 9.1 15.0 9.6 0.8 1.7 6.5 1.6 Synedra 2.5 11.1 1.5 3.B 3.7 1.7 7.4 2.3 18.2 0.4 Rhoicosphenia 2.6 2.5 11.1 2.9 1.2 0.9 2.4 1.0 2.7 0.4 Cymbella 11.1 0.7 0.8 1.7 1.2 O.B 1.0 0.3 1.3 0.4 Navicula O.B 0.9 0.5 Cyclotella 2.5 0.7 0.8 0.2 Stephanodiscus 0.4 0.2

N IJ1 W Percentage compositlon of the faecal material of Xanthocnemis zealandica, in terms of major food categorles, a projt~cted area basis (2 September 1976 - 21 November 1977 and oVerall for three size classes.

Size class 1 Size class 2

2 2 t~ov ~ 2 Dec. 20 Jan. Jun. OVERALL 2 Sep. 2 riov~ 2 Dec. 20 Jar.• 8 Apr. 20 Jun OVERALL

i9.7 18.6 17.5 14.E> 24 .. 2 .3 35.1 6 12,,5 15~1 21" 1

Detrl~US :L6 'L ~ 12.1 L4.9 .1 .1 9 .. 1 16 12.5 12 ~ 7 11,> 3

Macrophyte 7.6 J,9 ';.2 5.0 7 .1 .9 7.3 !.i.O 5.6

5,,\) Fi lamE:ntous alga.: 8.6 4.4 0.3 3.7 3.5 .4 6.1 J.

Anunal 52.6 64.u 61.2 54.7 58.5 45.4 65.0 56.0 58.7 5::'.9

1 • a.s [) 1 5.4 1~4 ".~

NUlnber of 1 arva€' 7 2 26 10 3 10 ')3

SiZe class 3

Se~ .. 2 Nov. 2 Dec 20 Jan. 8 Apr. 20 Jun. I Nov. Ov'ERl,LL

,. ~ Diatoms 1-' ~, 26~D 18.5 ILl f.4 12.6 , . ~w. /

Detritus 17.8 13.~ 12~4 19.1 6.9 12,,7 12 7 14.9

' ~ Mdcrophyte L8 2.4 5 ~ ., .i.~ 6.6 4.3 4.8 ., , Filame~~tous alga", ".J. 2.G 3. 3.5 9.2 7.4 4~S. 4~0 Animal 41.9 56.1 59.7 61 0 7E>.3 60.1 61.2 59.4 S.R.T' 5 O. -; 0.2 0.6 1 0.3

r~u...wer of larvae 2 8 10 ' r 8 exaroin~d J." 1 8 47 Appendix 6.9 Generic composl.tion (<0 by nwr.bers) of diatoms in the faeces of Xanthocnemis zealandica (2 Septerr.ber 197" ~ 21 November 1977) • O.h. (5ee Appendl.x 6.8 for nl..linber of larvae examined.)

Size class 1 Size class 2

2 Sep. :2 Nov. :2 Dec. 20 Jan. 20 J .. m. O\It:RALL 2 Sep. :2 Nov. :2 Dec. 20 Jan. a Apr. 20 Jun. OVERALL

COCCOlleis 53.4 91.2 83.0 6 69.9 72.4 51. 3 72.0 86.4 87.1 73.5 64.8 69.3 rragilaria 6.4 1,,6 4.0 .2 5.8 5rl 9.S 6,,4 5.5 3~4 10,,9 7.3

~ Epithem.la 6.8 0.9 4.5 4~1 ll~~ t.G 5 5~E. 4.1 • I 8~g S,8 1 S,:!nedra 17 .4 2.7 0 9 5 ~ 1.3 .1 4.6 0 3~4 :3.1 5.7 Asterionella 5.1 1.8 2.6 2.6 12.0 2.3 0.5 5.4 1.7 4.8 Cymbell~ 4.2 C~S .5 3.2 2.2 6.6 4><1 .7 0.5 1.4 3~1 Cy;;;;locel.J.a 0.9 1.9 .6 1.3 0.3 .7 2.5 5.8 1.6 Gomphonel7liil 4.2 2.7 2.2 O. 2 .. 4 1.6 2.8 0.7 3.2 1.5 2.7 1.7 Rho 1 cosph en J-ii 1 .. 3 0.9 6~1 ':'~G O~5 1.5 0.3 1.0 0 .. 6 Navicula 1.3 U","" 0.3 0.5 0.2 Melosirii!. 1.1 3 Pinnula.ri~ 0,.5 0.1 0.2 IJ) 0.2 T",J;;c )lilr ia 0.9 4.5 1.2 0.2

SiZE class 3 2 Sep 2 Nov. 2 20 Jan. 8 Apr. 20 Jun. 21 Noy. OVERALL

Cocconeis 5:>.9 66.3 82. 81.3 88.9 67.6 65.9 13.3 Fragilaria 3.2 8.0 4.8 2.8 5,9 6.2 5.2 Epichemia 14.0 4.4 6.5 5.7 ILl 7.1 7.,0 6.5 Synedrd 4.3 3.6 0.5 1. 1.r; 4.0 2.2 As ter.ionella 2.2 9.2 .5 3.2 2.4

~ Cymbella 10.8 5,,3 1.2 • I 0.4 2.1 Cyciotelld 1.1 0.3 4. 5.5 B.7 4.8 3.8 Gomphanel7liil 3.2 1.5 0.5 0.5 2,,4 6.2 1 .n ., Rhoicasphenia 0.9 1.2 O. 0.7 Navicula 1.1 0,,5 1.6 5.1 1.2 Melosira 4.3 0.3 0.3 n 1 Pinnuldria 0.3 0.2 ~ .. OpephoIa 1.0 1.2 0.4

N 1...r1 if> 256

Appendix 6.10 (a) Percentage composition of the faecal material of potamopyrgus antipodarum, in terms of major food categories, on a projected area basis; and

(b) generic composition (% by numbers) of the diatom category. * =: <0.1%.

(2 September 1976 ~ 21 November 1977)

1976 1977 2/9 2/11 2/12 20/1 S/4 20/6 21/11 Overall

(a) Diatoms 44.9 52.1 67.S 47.6 15.S 64.3 6S.6 50.S Detritus 19.0 21.5 lS.9 24.5 25.6 15.1 11.1 lS.S Macrophyte 24.1 16.0 5.2 19.3 36.2 12.2 16.S 19.7 Filamentous algae 11.9 10.4 7.S 6.0 22.4 4.1 3.6 9.5 S.R.T's 0.1 0.3 2.6 4.4 1.2

Number of snails 12 6 6 6 6 6 6 4S examined

(b) Rhoicosphenia 21.S 4.0 12.1 47.4 3.3 43.9 7S.1 39.0 Cocconeis 3S.5 63.4 50.S 36.S SO.3 37.3 14.0 3S.3 Epithemia 13.2 19.5 26.1 6.9 13.9 5.6 6.3 10.6 Fragilaria 2.0 1.0 1.0 O.S 9.6 O.S 3.2 Asterionella 9.S 1.5 4.5 0.3 O.S 2.5 Gomphonema 2.9 5.4 2.5 3.1 0.3 1.7 Synedra 6.3 1.5 1.5 1.6 0.3 1.6 Cyclotella 2.0 1.0 2.7 1.7 1.1 Melosira 4.S 0.8 1.1 Cymbella 1.6 0.5 1.4 0.5 0.6 Navicula 0.5 0.5 0.1 Diatoma 0.7 0.1 Opephora 0.3 * pinnularia 0.2 *